2.1
The automation of mass transit is a potentially revolutionary development in transport. However, it needs to be considered within the context of the planning of our cities and regions, and transport automation more generally. There are significant potential benefits in the automation of mass transit, but they will only be realised if we develop automation to meet the needs of our cities and regions.
2.2
This chapter will explore the wider context of automation before examining the benefits of transport automation. It will examine specific aspects of automated rail and road mass transit, before analysing the need to develop integrated transport systems joining mass transit systems to the wider transport network. It will then consider some of the implications of the transition to automation.
Context—planning
2.3
The Planning Institute of Australia (PIA) asserted that ‘the starting point for the Inquiry should be asking “what performance do we want of our cities?”—and thence: “how can various transport / energy /digital innovations contribute to achieving these goals?”’ The PIA was concerned that much of the current discussion around transport automation is ‘highlighting some mobility roles and benefits of connected and autonomous vehicles (AVs) in mass transit applications—but ignoring whether they help or hinder the achievement of broader liveability, accessibility and productivity expectations of our major cities and their communities into the future’. It urged an examination of automated transport within the context of ‘a national framework for infrastructure decision making and services investment based on a national vision for the performance of our cities and regions’, such as set out in the Committee’s recent report Building Up & Moving Out. The PIA sought ‘to establish a strategic context for considering the role and impact of various forms of autonomous vehicles’. It was concerned that ‘a forensic assessment of the transport task and the problems various AVs are intended to solve is lacking from the public discourse and is being lost in the hype generated by global vehicle makers’.
2.4
The Bus Industry Confederation (BIC) urged the pursuit of ‘more compact settlement patterns, through strategic land use development directions, which forms essentially a polycentric +corridors + neighbourhoods development model’. It argued for ‘spatially-oriented transport directions to support this land use direction’, including:
ensuring strong radial public transport to the centre, where capacity increases are required to cater for continuing strong growth (roads simply cannot carry the expected increased demands and have high external costs in/through central/inner areas)
improving circumferential arterial roads. Road-based public transport and freight should be prioritized in use of these roads
providing fast and frequent trunk public transport services supporting inner/middle urban nodes and development corridors, including for circumferential movement (particularly buses), linked to the cluster (node)/transit corridor development focus
better public transport connections from outer suburbs to areas of employment/activity concentration, including the small number of high tech knowledge-based clusters
upgraded trunk arterial roads in outer growth areas (to deal with the current backlog rather than encourage further sprawl)
increased local public transport opportunities in outer neighbourhoods, to support delivery of 20 minute neighbourhoods
improved walking and cycling opportunities throughout the whole city, with a particular focus on clusters/nodes and facilitating a city of 20-minute cities.
2.5
The BIC concluded that ‘the move to driverless vehicles needs to be understood in association with other transport changes taking place, such as shared vehicles and road pricing reform’. It argued that:
These matters must be treated in an integrated way, in the context of the broader societal trends and issues that have to be addressed – road safety, personal safety and security, national security, social inclusion, population growth, urban sprawl, possible job losses with technology, climate change and the need to address greenhouse gasses and traffic congestion.
2.6
In its study of the impacts of automated transport, Infrastructure Victoria identified significant implications with the new technologies for how cities might work. Infrastructure Victoria’s modelling demonstrated changes in where people were likely to live and work due to automation providing ‘greater accessibility in a transport sense to jobs and services’. As Dr Jonathan Spear, Executive Director and General Counsel at Infrastructure Victoria, noted, ‘where you have good access to roads as well as public transport networks’, those places are ’likely to become more attractive to live and work’.
Context—automation
2.7
The automation of mass transit must also be seen within the broader context of transport automation more generally. The automation of personal passenger vehicles will impact the development of mass transit.
2.8
Infrastructure Victoria has conducted ground-breaking research into the ‘impacts and potential outcomes of automated vehicles, vehicles on demand and zero emissions vehicles’. A range of scenarios was developed modelling ‘impacts on road congestion, the size and efficiency of the vehicle fleet, physical activity and access to activities and services’, and the ‘impacts of new vehicle technologies on the transport network as a whole, considering both public and private transport’. Infrastructure Victoria found that ‘under a future with automated on-demand vehicles, our findings point to a blurring of the traditional distinction between public and private transport as automated vehicles help supplement or fill gaps in the public transport network and improve access to services’.
2.9
Different scenarios had different outcomes for public transport usage: under the ‘Fleet Street’ scenario, the entire vehicle fleet would be electric, automated and on-demand by 2046, ‘with a cost per trip of approximately 30% of the cost of a current on-demand vehicle trip (such as an Uber)’:
Under this scenario, the total number of vehicles in Victoria is projected to fall by 93% to just 260,000. This is the result of shared vehicles being utilised for 36% of the day (compared with the base case where traditional vehicles are used just 4.8% of the day). Total vehicle kilometres travelled in this scenario are projected to fall by 15% due to a mode shift from cars to public transport as a result of the higher perceived cost of using on-demand vehicles. In total, 72% of trips were made by car versus 28% by public transport.
2.10
Under the ‘Private Drive’ scenario, all vehicles are ‘privately owned, electric and automated’:
In this scenario, private cars became even more ubiquitous, with 7% more cars than in the base case. Public transport use across all modes declined, falling from 19% of all trips in the base case to 14%. Buses were forecast to see the most significant fall in use (32%), but tram (28%) and train (22%) use also significantly declined. Despite this, overall public transport trips increased significantly compared to 2015 as a result of population growth, with nearly 3.7 million trips forecast in 2046 versus 1.7 million trips in 2015.
2.11
Under the ‘more likely future’ scenario, Infrastructure Victoria ‘considered a mix of vehicle types and ownership models that reflect the differing needs and preferences of users in different places’:
From modelling this future we found that road congestion could fall considerably (average delays are projected at 90% less than in the base case) and the size of the vehicle fleet could be significantly smaller than the base case (30% less than base case). This future is also forecast to lead to an increase in public transport use across all modes. Public transport use is forecast to account for 22% of all motorised trips, compared to 19% in the base case. Buses and trams in particular are forecast to experience the strongest growth in demand. This is likely due to public transport use becoming more common for more trip types, other than travel to work, and the fact that buses and trams service a more diverse range of destinations.
2.12
Infrastructure Victoria made ‘17 key recommendations which sought to help navigate through the inherent challenges and uncertainties that new technology brings, while maximising the benefits and minimising the risks associated with its introduction’. These were:
Integrate new transport options
Integrate transport management
Transition to zero emissions
Encourage demand management
Create planning flexibility
Lead and collaborate upgrades.
2.13
Within this matrix of recommendations, Infrastructure Victoria identified a number of ‘low-cost, no-regrets actions’ that governments could implement now. Infrastructure Victoria also identified significant investments that will be required to maximise the benefits of automation, including:
Up to $1.7 billion to upgrade mobile networks
Around $250 million for improved line markings on roads
At least $2.2 billion for energy network upgrades.
2.14
Infrastructure Victoria observed that there were many actions that governments could take ‘right now’ to prepare for new vehicle technologies:
These include integrating on-demand and mobility as a service offerings into the public transport mix, sharing transport data in real time and allowing flexibility in the planning regime to make it easier for people to charge their electric vehicles. All could have an immediate impact and deliver benefits, regardless of how new vehicle technologies roll out.
Benefits of automation
2.15
The benefits of automation were highlighted in a number of submissions. Infrastructure Victoria noted that its research ‘found that automated and zero emission vehicles could significantly reduce traffic congestion and greenhouse gas emissions, dramatically improve access to services, avoid car accidents caused by human error and add almost $15 billion per year to the economy’. The Department of Infrastructure, Regional Development and Cities (DIRDC) listed the benefits of automation as:
Safety
ensuring safer travel through being equipped with collision avoidance technologies; and
reducing personal safety risks, such as walking through unlit areas, by offering more convenient on-demand first and last mile transit options.
Efficiency
making existing services more cost-efficient for operators and passengers;
making new services more financially viable; and
reducing the need for investment in new services and infrastructure through network efficiency gains.
Accessibility
providing for new services in areas not linked by public transport and in areas of low patronage, particularly in regional centres;
providing increased accessibility to existing services by providing more convenient first and last mile transit; and
enabling greater mobility for people who may not currently be able to drive, such as people with a disability and older people.
2.16
DIRDC cited research indicating that ‘human error may be a factor in more than 90 per cent of crashes, and that road user distraction or inattention is a contributory factor in around 10-30 per cent of road crashes’. DIRDC noted that ‘if automated technology reduces or eliminates human errors, as is generally expected, then benefits for road safety may be substantial’. The safety benefits would also ‘extend to other vulnerable road users, such as pedestrians and cyclists, as vehicles with higher levels of automation will be able to detect their presence and take evasive action automatically’.
2.17
Nonetheless, DIRDC highlighted the potential difficulties of the transition to automation. It cautioned that while the safety benefits of automation were likely to be significant, ‘the effect of higher levels of vehicle automation on road safety remains untested at a large scale and may not be immediate or linear’. Moreover, ‘complexities may also arise from how automated and non-automated vehicles co-exist with potentially different driving behaviours’. DIRDC noted that ‘interactions between automated vehicles and vulnerable road users (primarily pedestrians and cyclists) pose potential risks, with scope for automated vehicles to fail to detect or accurately predict the behaviour of vulnerable road users’.
2.18
DIRDC also noted that ‘automated vehicles have the potential to reduce congestion and improve the efficiency and productivity of Australia’s transport networks by’:
increasing average traffic speeds and safely reducing following distances (headway) between vehicles;
optimising driving behaviours and routes, especially for trips involving multiple passengers with varying origin-destination needs;
providing increased reliability of travel times;
reducing stoppages and delays from traffic incidents;
encouraging the use of public transport through low-cost, on-demand first and last mile connections; and
facilitating more efficient movement of freight.
2.19
DIRDC observed that the ‘increased accessibility of transit services is likely to be particularly valuable in regional Australia’:
Road crashes disproportionately impact regional Australians; 65 per cent of road fatalities occur in regional, rural and remote Australia (BITRE, 2017c). Available evidence suggests that human factors, such as distraction and alcohol, are the primary causes of these crashes (Siskind et. al, 2011). Given that automation could reduce or remove the human element of driving, the potential to reduce crash rates is significant – particularly in regional Australia.
2.20
The emergence of automated vehicles would also provide potential for older Australians to continue to engage with and participate in the community. According to DIRDC:
In 2015, there were an estimated 3.5 million Australians aged 65 and over, representing one in seven people (15.1 per cent). Older people may have mobility challenges and as a result, they are less likely to be able to own a private vehicle or be able to drive.
As automated vehicles become more sophisticated over time, it is likely that there will be a reduced or removed expectation that the occupant will need to be ready to take control of the vehicle if required. Highly automated vehicles will enable older people to continue to visit the doctor, do their shopping and participate in the community (Siorokos, 2016).
2.21
There were significant potential benefits in vehicle automation for disabled Australians as well:
In 2015, almost one in five Australians reported living with a disability (18.3 per cent or 4.3 million people). 34 per cent of people with a disability report difficulties using public transport, with 14.7 per cent of people with a disability reporting inability to use any form of public transport (ABS, 2016).
Highly automated vehicles could mitigate these challenges by providing more convenient access to alternative transport options for people with a disability. Infrastructure Victoria (2018) has estimated that the economic benefit of increased transport accessibility created by the use of automated vehicles to be $3.5 billion per year in today’s terms. Trials of automated vehicles are increasingly considering the potential of automated technology to provide people with new mobility options.
2.22
DIRDC observed, however, that the benefits of automation would be dependent on how it was managed:
Increasing the comfort of in-vehicle journeys and reclaiming commuting time for leisure, work or sleep in highly or fully automated vehicles is an important priority for commuters and hence for vehicle manufacturers. However, this could reduce the perceived cost of travel time and provide an incentive for longer commutes, with a flow-on impact for land use planning and infrastructure provision.
2.23
Arup highlighted both the benefits and potential problems of automated transport:
Already, cities worldwide are experiencing an increased number of trips using “ride-share” services, claiming even more space. The ease of access has created more demand in combination with a perceived improvement in customer service and value for money.
Autonomy would significantly reduce the cost per kilometre, creating a more competitive market focused on providing customer convenience. That customer convenience would be at the externalised cost of significantly more vehicles operating empty, taking space away from more productive uses. Their contribution to congestion could be significant if unconstrained, as fleets of competing and autonomous Mobility as a Service (MaaS) vehicles rove the streets, being available for potential customers. Consequently, this type of transport needs to be minimised in inner cities and spatially constrained spaces.
2.24
The potential perils of shared mobility were highlighted by AECOM, who noted that ride sharing had actually increased congestion and seen lower public transport use in a number of markets. Mr Roger Jeffries, Technical Director, Transport Advisory and ANZ Technical Practice Leader, Transport Advisory at AECOM, explained:
New York City has seen a massive increase in congestion with the shared mobility market. What we've seen around the world with shared mobility—and we're talking about the likes of Uber and Lyft and other operators; it's not about any one operator; there's a whole range of operators—is that in the cities where there are advanced markets for those operators there has been an increase in congestion. What they've typically done is cannibalise the public transport market. As in the example of New York City, there were 30 million fewer subway trips in one year, related to the intervention of the ride-sharing market. That's not a market which is universally equitable to everyone in society; it tends to pick up the top 10 per cent of the public transport market. People who have the economic means to travel in a slightly more comfortable environment, one might say, for very little more—a marginal cost on top of what they might pay for public transport—can go in a private ride, or a near-private ride if it's shared with one or two other people, which will take them directly from point A to point B. Whereas if you ride the subway you may have to change once or twice, and that's maybe not quite as convenient.
2.25
Mr Jeffries stated that we ‘need to think about how those operators can be brought into the mobility mix—both to provide end-to-end journeys but also, potentially, to encourage them to supplement the mass transit network, as was alluded to earlier today—in a way that does actually support the city outcomes that we want rather than, as I said before, resulting in a perverse outcome for the city’. He noted that in New York,’ the state or city government has introduced restrictions on the number of rideshare vehicles that can operate within Manhattan’. He highlighted New York as an example of ‘issues that have been caused by a lack of regulation’, and suggested that Australia could ‘learn from things that have happened in a much more advanced market’.
2.26
Arup noted that the ‘autonomy of small passenger vehicles appears to bring the largest opportunities in less densely populated areas’, a particular issue given Australia’s sprawling population:
Not only because space is a more widely available resource here, but also because of large potential cost savings. Public transport is relatively expensive to deliver in suburbs, where revenue is small and bus drivers’ salaries make up the majority of operational costs. Driverless bus services could dramatically decrease public transport costs, creating opportunities for shorter local routes, less waiting time, distributed cross-regional connectivity, increased service frequency, more reliable networks and thus the promotion of more sustainable travel behaviour.
2.27
Monash University observed that ‘autonomous vehicles are projected to help alleviate congestion by reducing the overall number of vehicles on roads and parking spaces’. It suggested that ‘opportunities for speeding up traffic arise from autonomous vehicles to moving in platoons, optimising routes ad hoc and navigating intersections without traffic lights, especially if human driving can be made obsolete’. It noted, however, that ‘the uptake of autonomous vehicles by people not currently driving, empty trips, autonomous vehicles running errands and people currently using public transport is likely to increase traffic loads’. Monash found that fully automated self-driving point-to-point transport will increase the number of kilometres driven for the following reasons:
Additional trips will be made by people unable to use existing transport
Empty relocation trips have to be made
The absence of a driver makes travel more affordable
Removing the strain of driving frees up travellers to complete tasks during trips and makes longer trips more bearable
Autonomous vehicles may also be used to run errands without passengers.
2.28
According to Monash, ‘for point-to-point autonomous vehicle services to deliver on reducing congestion, trips have to be shared’:
Simulations by the International Transport Forum (2016) of 8-seat and 16-seat on-demand minibuses demonstrate that a reduction in traffic loads is only possible if more than 60% of all private vehicles are replaced by the shared mode. Declining car ownership and the resulting reduction in the need for parking space has no impact on the total travel distance but can improve traffic flow (Rantasila, 2015).
Gruel and Stanford's interviews also suggest that increased efficiency of operation and the opportunity to share autonomous vehicles in addition to using them as a feeder service will help reduce congestion (Gruel and Stanford, 2016). Alessandrini et al. report findings from European studies on urban mobility, most notably the realisation that shared fixed-route 4-seater autonomous vehicles are only likely to be effective inside local and city centres. For most other transits between different service centres, inner and outer suburbs, individual use autonomous vehicles are likely to be effective, as are fixed-route bus services.
In contemporary Australian cities, many people travel from the suburbs to the city for work and therefore share a large part of their routes to work with other people. Diversity of trip demands is likely greatest at the start and end of the journeys, suggesting fewer and smaller vehicles are needed in these parts.
Rather than covering the joint part in between start and end with large point-to-point autonomous vehicles, it appears meaningful to retain fixed-route transport (bus and train-like services) along major corridors, with small autonomous vehicles as feeder services that cover the first and last miles (Kelly et al., 2015).
Alternatively, very small ‘stackable’ individual-use vehicles that can be assembled into larger 'trains' are a conceivable if radical and expensive option. Travellers are averse to mid-trip mode changes, and this option would eliminate a need for relocation and possible wait times.
2.30
The Bus Industry Confederation identified two potential scenarios surrounding transport automation. In the optimistic scenario, ‘cheap, accessible, low/zero emission driverless vehicles are widely available on-call, either for single use or shared use but shared use predominates’. The ‘availability, convenience and cost of accessing AEVs [Automated Electric Vehicles] are such that people see less need to own their own vehicles’. BIC concluded:
The cheaper cost of AEV travel, particularly by ride-sharing, and the opportunity for new vehicular trips by mobility/transport disadvantaged people will combine to mean that the number of person trips increases in the optimistic scenario. Given sufficient penetration of shared mobility choices, however, this higher number of person trips can be satisfied with a slower growth in vehicle kilometres travelled, even though autonomous shared vehicles need re-positioning movements.
2.31
Under the pessimistic scenario, BIC indicated that:
… the personal appeal of private ownership, reinforced by the perceived lower cost of AEVs and opportunity to use travel time productively lead to increased personal trips, with vehicle kilometres increasing at least as fast as personal trips but most probably much faster, as car owners avail themselves of the opportunity to call up their car when they want it …
The opportunity to work-in-vehicle, or rest/sleep while travelling, instead of having to deal with the driving task, will be seen by some people as an opportunity to change place of residence, most likely to consume additional space by moving to the peri-urban area or even beyond, extending urban sprawl.
2.33
BIC saw ‘the consequences of greater urban sprawl as potentially the biggest single risk from widespread adoption’ of automated vehicles. It argued for an outcome based on the ‘substantial penetration of shared mobility’, indicating that success would be achieved ‘in part because policy settings explicitly target this outcome’. The BIC argued:
With vehicle use in the optimistic scenario now paid for on a more direct pay-by-use basis, active transport is likely to account for a higher mode share than in the pessimistic scenario, with multiple societal benefits (e.g., improved health, lower congestion). The higher mode shares for active travel will, in turn, be supportive of more compact settlement patterns than in the pessimistic scenario. One implication is likely to be relatively higher urban productivity from clustering in the optimistic scenario. Also, the more compact urban form will mean a lower level of infrastructure spend on the urban fringe and beyond, easing government borrowing requirements.
2.34
The BIC argued that:
The introduction of driverless vehicles should be seen as an opportunity to review mobility in general, reflecting on the whole mobility system, the purpose and value of mobility and how it can be accomplished better in social, environmental and economic terms, recognising the potential benefits and challenges associated with driverless vehicles.
2.35
It believed that ‘mass transit in the future could be very different depending on the policy setting of Governments’.
Automated rail mass transit
2.36
Train automation ‘refers to the process by which responsibility for operational management of a train transfers from the driver to the train control system’. Four grades of automation (GoA) are recognised internationally:
GoA1—Automatic Train Protection (ATP) with driver—driver controls starting and stopping of train, door closure and operation in event of disruption
GoA2—ATP and Automatic Train Operation (ATO) with driver—starting and stopping of train is automatic; driver controls door closure and operation in event of disruption
GoA3—driverless—starting and stopping of train is automatic; train attendant controls door closure and operation in event of disruption
GoA4—Unattended Train Operation (UTO)—starting and stopping of train, door closure and operation in event of disruption is automatic.
2.37
The key elements for automated rail systems are:
Automatic Train Protection (ATP) is a system designed to avoid collisions, and help prevent red signal overrunning and exceeding of speed limits by applying brakes automatically. A line equipped with ATP corresponds (at least) to a GoA1.
Automatic Train Operation (ATO) insures partial or complete automatic train piloting and driverless functionalities. The ATO system performs all functions of the driver, except for door closing. The driver only needs to close the doors, and if the way is clear, the train will automatically proceed to the next station. This corresponds to a GoA2. Many newer systems are completely computer controlled, while still electing to maintain a driver or a train attendant of some kind to mitigate risks associated with failures or emergencies. This corresponds to a GoA3.
Automatic Train Control (ATC) automatically performs normal signaller operations such as route setting and train regulation. The ATO and the ATC systems work together to maintain a train within a defined tolerance of its timetable. The combined system will marginally adjust operating parameters, such as the ratio of power to cost when moving and station dwell time, in order to bring the train back to the timetable slot defined for it. There is no driver, and no staff assigned to accompany the train, corresponding to a GoA4. This Grade of Automation is also referred to as Unattended Train Operation (UTO).
2.38
Automation in the rail sector is not new. The development of automated systems and technology has been going on for decades and continues apace. DIRDC notes that ‘in the rail sector, automated transport is a proven technology, giving rail a head start over other land transport modes’:
Fully automated, driverless rail networks have been in operation for over 30 years, with the most advanced rail technology available today allowing for unattended train operation with no staff on board. As of May 2018, there were 63 fully automated operational metro rail lines in 42 cities in 19 countries across the world, including major cities such as London, Paris, Lille and Singapore. The total line length of fully automated metros reached the milestone of 1,000km in 2018. Over the period 2015 –17, ten new metro lines designed to run with fully automated operation entered in service in ten cities (UITP, 2018).
2.39
DIRDC observed that ‘automated train technology continues to grow at a rapid rate. Current forecasts, based on confirmed projects, indicate that by 2025, there will be over 2,300 kilometres of driverless metro lines in operation worldwide’.
2.40
DIRDC highlighted the benefits of automation in rail, stating:
Within dense urban environments where new rail lines can have high upfront capital costs, technological advances and increased automation have the potential to improve the efficiency, capacity and utilisation of existing networks to help cater for expected growth in patronage and attract more people onto rail through customer focused initiatives.
2.41
The ways in which automation can transform the use of mass transit include:
more trains, more often – automated rail systems allow operators to optimise the running of trains, increasing the average speed of the system, shortening headways and reducing dwell time in stations;
greater reliability – driverless technology and advances in communication and control systems increase resilience by allowing dynamic, real time management of the network in the face of disruptions and enabling operators to allocate trains in response to sudden surges in demand. Automated, independent lines also help to improve the reliability of overall networks through untangling networks and better geographically containing network-wide delays;
safety and accessibility – automated rail systems offer safer operations than conventional railways by reducing the human-risk factors and increasing reliability (UITP, 2011). Upgrading to a more modern system design also brings a number of added safety and accessibility benefits. For example, newer metro systems can be built to facilitate level access between platform and train, and platform screen doors can prevent passengers tripping or falling onto the tracks;
personalised journeys – technologies, including automation, create opportunities to provide customers with real time information and facilitate the use of public transport through on-demand first and last mile connections.
improved energy performance – acceleration and deceleration patterns of automated rail are adjustable to reduce energy consumption and maximise energy recovery, therefore significantly reducing energy costs; and
reduced operational costs – communications-based train control systems allow for the removal of traditional trackside infrastructure, such as track circuits and colour signals, and the associated costs of maintenance.
2.42
DIRDC stated that as a ‘step towards greater automation, Automatic Train Protection (ATP) is increasingly being rolled out across Australian rail networks’. It noted that ‘in NSW, for example, ATP is being rolled out across the Sydney Trains and NSW TrainLink electrified network’. ATP is used to monitor train speed, distance and direction and ‘prevents the authorised line speed being breached due to driver error or if a driver becomes incapacitated’.
2.43
DIRDC highlighted the development of automated Metro as a significant development in automated rail systems:
Australia’s urban rail systems will undertake a step-change over the coming years with the completion of Stage 1 of Sydney Metro – the NSW Government’s $8.3 billion Sydney Metro Northwest railway. Sydney Metro is an automated mass transit solution that will bring congestion busting and city-shaping benefits to Sydney, transform urban centres, and boost economic productivity by improving employment and education opportunities.
2.44
It observed that the Sydney Metro Northwest project ‘will be fully automated with no train attendants present’, with ‘controllers monitoring the entire system from an Operations Control Centre’. Nonetheless, safety and security was central to its design:
The Sydney Metro Northwest trains are being designed, built and operated to the highest safety standards, with more than 300 Australian and international safety standards stipulated in the operations contract for the trains and the associated equipment. High levels of security will prevent trespasser access, such as platform screen doors that keep people and objects away from the tracks and allow trains to enter and depart stations faster. Obstruction detectors will prevent trains departing stations if any door is not fully closed. An intrusion detection system will be a feature on Sydney Metro Northwest, designed to identify and report any track encroachments along the route.
2.45
DIRDC further noted that ‘Metro rail systems are often closed systems with access restricted to the automated vehicles and controlled access to other types of trains, road vehicles and pedestrians. This element of system design is linked to safety outcomes’.
2.46
The Australasian Railway Association (ARA) believed that the Sydney Metro would provide a model for future passenger rail development, illustrating ‘what is possible for automated mass transit systems in Australia’ and paving the way for ‘additional automated, driverless rail lines in Australia’. It noted that ‘projects such as the Melbourne Airport Link are also considering driverless automated trains’.
2.47
The ARA highlighted other important advances in automation including:
Digital Signalling Systems / Automatic Train Control—which ‘removes the reliance on track-side signalling, using automated systems that allow passenger and freight operators to provide more services using existing infrastructure as trains can be run closer together’, thereby increasing ‘network capacity and improve the customer service offering by reducing wait times for customers and helping to manage station crowding’;
Automated asset monitoring for maintenance—informs and automates ‘maintenance decision making to maximise the use of high cost assets by reducing the impact and lost revenue of unavailable rolling stock or infrastructure. This smarter approach to monitoring and asset maintenance can also extend the lifecycle of parts with decision-making based on data analytics and insights.’
Smart Ticketing—which provides ‘many benefits including streamlined customer travel experience; negating the need for single, weekly or monthly use paper tickets; providing extensive customer travel insights for operators and Government alike, ensuring customers pay correct travel fares’; it is expected that the current systems will be succeeded ‘by mobile phone or credit card ticketing systems’ improving ‘the accessibility of and integration of our transport systems’.
iTRACE—an initiative ‘implementing global data standards (GS1) in the rail industry to standardise the way all assets and materials in the rail industry are identified, barcoded and tagged’. Standardising the capture of data ‘ will help to improve efficiency, lower costs and deliver better customer service and bring industry-wide efficiencies by setting the foundation for automation’.
2.48
The ARA cited the industry’s Smart Rail Route Map as a reflection of the ‘shared desire to identify a long-term vision for technology in the rail sector through the establishment of a common view of priorities, themes, timelines and actions for the next 30 years’. It focused on the opportunities and challenges associated with ever increasing levels of automation in the Australasian railway industry, including:
New technologies requiring a shift in the skill-sets in the rail sector towards automation, and that people management skills will move towards the interface between human and digital workplaces.
Automation paving the way for greater simplification of journey planning, allowing greater access to a reliable, multimodal transport service, with improved last mile connections.
Potential opportunities existing for greater integration of cloud-based computing, analytics and other systems to enable automation of traffic and network management systems.
Complex systems will be simplified through automation, balancing capacity and flexibility, while humans will maintain supervisory control of technology.
Developing an industry platform to monitor where the industry is heading with regards to automation, why the direction is important and the priority tasks, helping the workforce visualise the opportunities and potential pathways to the future.
New systems, based on automation, allowing traffic management to progress to a role of train service optimisation and significantly improve track capacity and train safety.
2.49
The ARA cautioned, however, that achieving ‘interoperability between automated systems, particularly across State borders is integral to ensure the rail industry does not end up with another “break of gauge”’.
2.50
Another concern was raised by Monash University, which noted that ‘the quality and age of Australia’s rail and bus infrastructure’ was a barrier to automation—‘the larger Australian cities have very old legacy railways and unreliable infrastructure. This will have to be renewed to make automation possible.’ Nonetheless, Monash believed automated rail transport was the way of the future:
Automated driverless trains are becoming common in various countries. In Australia, the Office of the National Rail Safety Regulator recently approved Rio Tinto’s auto haul of 240 car heavy haul cars. There are a number of opportunities for Australia to lead in the development and implementation of new technologies that will assist operators where automated driverless trains are used, create greater efficiencies and improve safety. The scope includes light rail, passenger rail networks and in the heavy haul freight rail industry.
2.51
In its submission, Arup stated that ‘with more than 30 years of technology experience it is beyond debate that autonomous rail mass transit is safer, more reliable, cheaper to operate and can deliver more services per hour on the same infrastructure footprint, made possible by segregated infrastructure’. It urged the adoption of Automatic Train Operation and Automatic Train Protection in ‘all major city passenger rail networks in Australia as a step to full autonomy’. Arup noted that ‘autonomy of rail-based freight transport is not as far progressed as rail-based passenger transport, but has potential of equal measure’, stating:
The question where in the freight chain autonomy can best be applied is still largely unanswered and needs further investigation. Options include autonomy on trunk routes, terminal transhipment or last mile delivery.
2.52
Arup observed that ‘it is most likely that those parts that can be operated using electric motivation will become autonomous first’. It noted that ‘electric powered movement increases predictability, allows for smooth acceleration and deceleration and is easy in use, therefore lowering costs of automation and operation’.
2.53
The Bus Industry Confederation made s further observation about rail mass transit—the need to provide strong government control over a natural monopoly:
The high capital costs and associated high patronage of rail mass transit services to central cities provides them with significant natural monopoly characteristics, which suggests multiple sources of supply are unlikely. The agglomeration economies, congestion cost savings and environmental benefits (external benefits) associated with such services speak to the importance of strong governmental control over service provision, rather than leaving them to the dictates of the private marketplace, where under-provision would be expected, relative to the scale of external benefits. We conclude that these natural monopoly characteristics and external benefits are such that, in coming years, the Australian mass (trunk) transit market should remain as public transport as we currently understand it. There is a need to include these trunk services in MaaS bundles, for which they will provide a fundamental ingredient.
Rail v. Road
2.54
Despite the convenience and benefits of automated cars, there was a widespread belief that rail mass transit would continue to play an essential role in providing transport solutions. The ARA observed that while ‘automated, driverless cars provide an exciting revolution in transport and are a clear potential to provide the “first and last mile” for public transport, mass transport such as heavy and light rail will still be vital to provide the spines for seamless integrated mass transit systems’.
2.55
DIRDC observed that the mathematics of mass transit meant that rail would always be more efficient than road on trunk routes, stating:
On an arterial freeway, you’re looking at a little over 3,000 people per hour. If you’re talking about, say, a high-capacity metro train, it’s more like 34,000. With the Sydney Metro, it’s potentially over 40,000 people per hour. So, even if you do get a substantial improvement in your carrying capacity on a road because of autonomous vehicles, it’s pretty unlikely that they’re going to get to that sort of level, because that’s an order of magnitude more people being shifted on heavy rail options, for example.
2.56
The ARA noted that ‘Australia’s growing, aging and urbanised population is putting increasing pressure on our public transport systems, of which rail provides the backbone, moving the masses’. It stated that ‘governments and rail organisations in Australia and around the globe are increasingly looking to automation to safely increase the capacity of existing infrastructure, improve the customer experience and ensure rail services modernise to meet the needs of Smart Cities of the future’. The ARA argued that ‘more trains and more services are required today and into the future to meet the needs of our growing population to travel and move freight’. Automation would ‘assist to provide the greater capacity required by passenger and freight rail networks’.
2.57
In its analysis, Monash University observed that ‘metro rail systems have been the backbone for most smart cities’. It argued that while ‘a bus can take up to 60 passengers’, to replace it ‘with cars, autonomous or not, it would require additional space on our roads’. But, ‘if you substitute cars with a metro rail system, it will be significantly more efficient and more effective in reducing congestion’.
2.58
Monash University also observed that ‘there are some critical lessons to be learned about the design and management of passengers in driverless urban railways which are important to note for running driverless vehicles on streets’. These included:
Driverless trains adopt platform doors to meticulously manage human interaction with entry/exit to vehicles
Rail platforms are generally underground or raised; and platforms do not permit any other vehicle of pedestrian interaction with trains
Streets where driverless buses or cars might operate have none of these protections and are far more complex locations; it is thus far more difficult to operate buses/cars without drivers safely in these places without a considerable degree of management of passenger interaction.
2.59
Engineers Australia urged priority for rail automation over road automation, stating:
A legal and regulatory framework exists for automated rail, so migration to driverless vehicles in closed systems such as rail networks should be prioritised. Full automation of our rail transport networks may be achievable sooner than automated road mass transit and public risk perception towards driverless vehicles may be tempered by a rail first approach.
Light rail
2.60
DIRDC noted the expansion of light rail in Australia’s major cities. It observed that ‘modern light rail systems are expected to increasingly adopt automation technologies that improve network safety and navigation’, but that as light rail ‘interacts with road traffic and pedestrians, it is less suitable for driverless automation than heavy rail’. DIRDC highlighted ‘recent technological developments in road-based mass transit options include trackless trams’:
Trackless tram technology varies, but generally involves a tram-style vehicle with rubber tyres that runs on markings on the road surface with the capacity for high levels of automation. Advantages of trackless trams, compared with traditional light rail, is that they can be significantly cheaper to deploy due to lower upfront infrastructure costs and much lower impact on the community during the construction process.
Automated road mass transit
2.61
There are six levels of road vehicle automation as defined by the Society of Automotive Engineers’ (SAE) International Standard J3016. The classification is based in whether the system:
manages steering, acceleration and braking on a sustained basis;
requires a human driver to monitor the driving environment and respond as needed;
can operate without handing over control (‘falling back’) to a human driver ; and
can operate in all situations (‘driving modes’).
2.62
The six levels of automation are:
No automation (SAE Level 0)―human driver undertakes all aspects of the driving task.
Driver assistance (SAE Level 1)―in some circumstances the system is capable of either steering or acceleration/deceleration (including braking), with the expectation that the human driver performs all remaining aspects of the driving task.
Partial automation (SAE Level 2)―in some circumstances the system is capable of both steering and acceleration/deceleration. The human driver must monitor the driving environment and respond as needed.
Conditional automation (SAE Level 3)―Level 2, but when the system is operating in automated mode the human driver is not required to monitor the driving environment. The human driver must respond to requests from the driving system to intervene.
Highly automated (SAE Level 4)―Level 3, but no human monitoring or intervention is required when the system is operating in automated mode.
Fully automated (SAE Level 5)―automated system in control all of the time, and in all road environments.
2.63
DIRDC observed that ‘the extent to which a vehicle is automated, and in particular, whether a human is required to monitor the road environment and/or be ready to take back control, has significant implications for the social, policy and regulatory impacts’.
2.64
In its submission, NRMA highlighted timelines for the introduction of automated vehicles developed by the Australian Driverless Vehicle Initiative (ADVI), stating:
With Level 3 technology already embedded in some light passenger vehicles, ADVI expects the arrival of Level 4 technology between 2020 and 2025, and Level 5 technology between 2026 and 2030. This view aligns with timeframes previously submitted by the NRMA in public papers, including The Future of Car Ownership (August 2017).
2.65
NRMA stated that ‘fully automated vehicle capability or “Level 5” automation—where no human driver is needed and vehicles do not possess a steering wheel or pedals—could be available as early as the mid-2020s. On-demand shuttles and taxis capable of full automation could arrive even earlier.’
2.66
Infrastructure consultants Arup observed that the ‘increased productivity of road mass transit, including passenger buses, forms the largest and most important opportunity in Australia, due to the extensiveness of Australia’s road network and its low productivity’. It indicated that ‘automation of road mass transit can provide this increase in productivity, made possible by potential reductions of operational costs and safety improvements’. To date, however, automated buses had enjoyed limited success:
Autonomy of buses has been hampered by high costs, low operating life, a small supply chain of electric buses and safety risks. Existing autonomous buses are operating in very low speed, highly controlled environments. Even those to be introduced shortly and touted as operating on the public road network have significantly lower operating speeds, much higher quality of infrastructure and constant human oversight. However, as autonomy improves and sensor technology combined with machine learning become more meaningful, the potential is significant.
2.67
The Bus Industry Confederation also noted the limitations of automated buses to date:
The BIC would note however that the concept of a driverless bus, in particular large buses, may be technologically possible but the reality of mass transit and school bus services operating in this way are much less certain for a variety of operational and personal safety and societal issues. The unknown element from a bus perspective is if it is going to be accepted by users concerned about safety and security.
2.68
The BIC cited trials of automated buses in France, where buses operating with only passengers raised a number of concerns:
One factor that has been recognised after actual trials of driverless buses on guided busways in France is that passengers do have concerns of trust and safety when a driver is not aboard. In this example, drivers were returned to the bus to ease concern, despite the fact that the vehicle remained self-driven. The physical presence of the driver was an important psychological factor, even if it was only for “override” capabilities if required. Trusting future technology will be a major challenge for many individuals.
2.69
Nonetheless, the BIC saw automated buses as the key to Australia’s urban transport future. Automated buses, ‘operating on bus priority infrastructure and dedicated bus rapid transit infrastructure such as the Brisbane Busways’, would lead mass transit services. They would ‘become the train of the future in an autonomous world where fixed infrastructure is no longer required’:
A train set of bus seats travelling along a transport corridor where individual bus carriages have the capacity to peel off as required to deliver passengers as close to their end destination as possible and connect to on demand services to complete the trip.
2.70
This was Mobility as a Service (MaaS—see below), with ‘a spine of mass transit that carries the bulk of the population most of the time’. The BIC concluded that the ‘bus is the workhorse of Australia’s mass transit systems today, carrying more passengers than rail each day and this will continue to be the case in an autonomous transport world’.
2.71
In its submission, DIRDC highlighted the various trials that have taken place with automated shuttle buses. The Department stated:
The automated shuttle bus trials currently underway in Australia are demonstrating the longer term potential of highly and fully automated driving systems. Partially automated buses, which are likely to be available sooner, may also benefit bus drivers, operators and passengers.
Mobility as a Service
2.72
Mobility as a Service ‘is a framework which aggregates infrastructure, services, technology and information to suit the travel and lifestyle needs of individuals’. It ‘brings together transport operators and third parties, allowing a seamless provision of services, information, booking, payment and customer relationship management between transport modes’. Engineers Australia notes that MaaS is ‘an emerging concept’, the definition of which ‘is not yet universal’.
2.73
MaaS has been identified as a key element of an integrated transport network in an automated transport future. Infrastructure Victoria observed that ‘on-demand public transport and mobility as a service (MaaS) solutions could make significant improvements to how we travel’. Infrastructure Victoria ‘found that on-demand, MaaS and integrated planning and payment for multi-modal trips are likely to supplement existing public transport services and pave the way for introducing on-demand automated vehicles’. It recommended ‘incorporating on-demand and MaaS into the public transport mix in preparation for automation, through the following actions’:
Ensuring new contracts for public transport operators allow for changes to accommodate new market models.
Plan for opportunities to develop open payment, ticketing, validation, third-party purchasing platform(s) and open/integrated barrier systems for public transport.
Plan for changes to public transport hubs to accommodate pick-up and drop-off facilities, and other mobility options like active transport to encourage multi-modal trips.
Review existing contracts and public transport franchise agreements, including fare structure, for opportunities to integrate automated vehicles into service planning.
Transport services delivered directly by the government (such as community transport) should plan for potential changes to accommodate new market tools (for example, apps and on demand services) and automated vehicles.
Assess potential for automated vehicles to support demand-responsive transport services.
Consider whether there is a role for government to incentivise or procure services from automated fleet operators to operate in regional and rural areas, if the market fails to do so.
Consider how automated vehicles could be used to enhance public transport, especially to support people with mobility impairments and those currently on concession arrangements.
2.74
Infrastructure Victoria’s findings pointed to ‘a blurring of the distinction between public and private transport’:
In a future with automated vehicles, having simple and efficient interactions between private operators and public transport will be critical to unlock accessibility and community benefits.
2.75
Dr Jonathan Spear, Executive Director and General Counsel at Infrastructure Victoria, noted that their modelling ‘shows that automated on-demand vehicles can be particularly complementary to public transport, because they have the opportunity to supplement and fill gaps in the public transport network and improve accessibility’. He continued:
If the fleet is made up of a mix of public transport vehicles in a traditional sense and different automated vehicle types that are both publicly and privately owned and deployed, then there's a possibility to cater for the different needs and preferences of different users.
2.76
Dr Spear observed that while ‘road based automated vehicles that are smaller than buses are unlikely to replace mass transit’, they could ‘help fill in some of those gaps of services that you do see beyond the CBD and some of our activity centres, and may well help people get to those activity centres and agglomerate there’.
2.77
The Bus Industry Confederation pointed to developments in the integration of public transport (PT) operations and MaaS as a key development for the future of automated transport, stating:
Whilst the suggested future of PT operators absorbing the MaaS broking role within their business model constitutes a longer term development, much innovation is already happening, with forays into intermediate modes and new models of providing local (coverage) transit. Whilst this is evident from the innovative work of multinational multimodal operators (e.g., Transdev, Keolis) in overseas markets, local Australian operators are also keenly exploring this space. In NSW, on demand services have being trialled since late 2017 in the form of government-led pilots, with various models deployed in Metropolitan Sydney, Outer Metropolitan Sydney and (from late 2018) in Rural and Regional NSW. Existing PT operators are partnering with technology providers to deliver these new innovative services.
2.78
Within this scenario, ensuring universal access to the transport network was a priority. According to the BIC:
If social inclusion is seen as a societal priority, then some base level of shared mobility service to support or underwrite this outcome is warranted. We see no other way of assuring minimum local mobility opportunities are available to ‘at risk’ people. By implication, local shared mobility contracts should be developed to support provision of base social transit service levels, which would be expected to vary by demographic/land use setting. For example, expectations should realistically be for a lesser service level in a rural area than in a town.
2.79
The BIC preferred ‘a subsidized minimum service level approach to shared mobility service (social transit), which supports individual capabilities and allows people to self-select on use, with existing fare concessions continuing’. It stressed that ‘the subsidy for shared mobility service should be for service , not modes per se, and shared mobility contracts should reflect this focus’. BIC stated that ‘shared mobility contracts are most relevant in rural, regional and outer urban settings, where they could be introduced now, given sufficient institutional will to pursue more integrated service offerings’. It suggested that ‘in a time of “disruption”, the way we do mobility, the way we move people is changing, and a shared mobility future based around actual demand rather than latent demand is with us’. The BIC concluded:
At State and Territory level, early development and implementation of service-focussed shared mobility contracts would be a positive supportive step along the transition pathway to future governance models that are better suited to emerging technological opportunities, while delivering immediate benefits from realizing a more integrated service delivery model. Integrated app-based booking/ticketing systems, with a range of on-demand service options, are fundamental to the prospects for MaaS and for shared mobility service in the immediate future and should be a requirement of shared mobility contracts.
2.80
The Department of Infrastructure, Regional Development and Cities held a similar view. It stated:
Public transport is moving away from a means of simply commuting along fixed routes. A convergence between automated driving technology and emerging data and connectivity-driven technologies may support more cost-efficient, on-demand on-road mass transit services. On-demand services would offer flexible routing and timing to better meet commercial and passenger needs. New business models and digital platforms, such as Mobility-as-a-Service (MaaS) are likely to provide real-time information about demand to transport operators, enabling the more efficient provision of services.
2.81
The Department observed that ‘on-demand services using small or medium size automated buses with lower operating costs could significantly improve service coverage, including in both urban and regional centres’. It indicated that ‘this type of automated transport could be cost competitive with regional rail links or traditional bus services, or could fill last-mile service gaps’. The Department noted that New South Wales was conducting trials of on-demand bus services—allowing people ‘to book a vehicle via an online app or over the phone, representing a shift away from the concept of fixed transport routes’. The trial, ‘rolled out in locations such as Eastern Suburbs, Manly, Northern Beaches, Woy Woy and the Illawarra’, had seen over 150 000 customer trips taken in its first two years. The Department observed that:
The cost of on-demand services through the pilot program is comparable to a one-way bus ticket and it allows people to access public transport when and where they need it, enhancing the existing public transport network. This trial is an example of how smart planning and applying technology in an innovative way to transform existing services can not only make delivering transport more efficient for operators but also provide better services for the community.
The First & Last Mile
2.82
One of the biggest challenges for automated mass transit is the problem of the first and last mile—the gap between mass transit services and home or destination. Mr Ian Christensen, Managing Director of iMove Australia, told the Committee:
We would also highlight that automation, while important in its own right, delivers substantially greater benefits if it's augmented by considerations of connectivity at the same time—'connectivity' meaning that one automated vehicle can interact effectively with both automated and non-automated vehicles in its surroundings and with which it interconnects in the transport systems of which it's a part. For instance, we can think of mass transport services as arteries in the transport network. But, like any organism, arteries only work well if the contents can actually get to the arteries. So we would say that mass transport and automation of mass transport is good, but it needs to be augmented by interconnection or interoperability with the distribution services—the 'last mile' services, so to speak—for the people who are using those services. Otherwise, we might have wonderful mass transport systems that run sub-optimally, or which are in fact potentially empty because people cannot get to them.
2.83
Mr Christensen observed that ‘across the whole of the personal transport system there is a progressive trend and need to transition from a modal focus—that is, the train, the tram or the bus—to a traveller focus, so that we concentrate on the experience of the traveller rather than just the efficiency or frequency of the transport service’. He believed that the main imperative is ‘to mitigate congestion, to reduce the burden of congestion on our productivity’. To do this, ‘we actually then need to encourage a better spread of transport demand across the available services and capacities’. This required behavioural change on the part of the travelling public, which could only be achieved ‘when travellers perceive the alternative modes of behaviour as being attractive’. There was ‘an absolute requirement that we migrate to a traveller-centric focus on the performance of our transport network, with a view to encouraging behaviour change to the extent necessary to reduce congestion’.
2.84
Mr Christensen argued for a shift away from a focus on modes of transport to a ‘systems approach’, providing ‘appropriate transport services to a geography, to a region, to a locality in which the bus operator provides part of the solution’. He noted that this would require ‘an evolution, at least in some jurisdictions, of the nature of the contract between the state and the operator as to the performance they're required to deliver and what they're trying to optimise’. He noted that ‘in some jurisdictions there is no incentive to the operator to increase patronage and yet, overall, the system would benefit strongly, up to a limit, if the patronage on public transport were actually able to be increased or people could be attracted from their single occupancy cars into multiple occupancy public transport vehicles’.
2.85
DIRDC noted that the first/last mile ‘is often disproportionately inconvenient when compared to other parts of the journey and may have a negative impact on public transport patronage – frequently it is too far to walk, too close to drive (and find parking)’. Increasingly, mass transit users are ‘using ride-hailing and bike-sharing services for the first and last mile of their journey’. The Department cited figures that indicated ‘15 per cent of Uber rides in Western Sydney are to or from train stations, with this figure reaching 25 per cent in some cities in the United States’. It noted that:
Small and medium size automated vehicles could provide more cost-efficient and convenient first and last mile services to and from public transport hubs, or connect people to and from regional heavy rail services. Trials of automated vehicles are increasingly testing this proposition by deploying automated shuttle buses on fixed routes between mass transit hubs and key origin/destination points.
2.86
MaaS, combining ‘public and private transport options in a single app’, combing integrated planning, booking and payment options, was seen as the way forward. The Department noted:
New business models such as these could accelerate long-term trends away from car ownership, and impact on travel patterns and infrastructure use. If this model matures in Australia, it could provide an incentive for travellers to move away from private vehicle ownership and make increased use of automated vehicles and public transport as part of a new, flexible approach to travel.
Our cities need to be well connected and integrated to support changing consumer expectations of rapid, just-in-time and distributed mobility. Australia, with its high concentration of urban populations and long, narrow linkages between them, is well placed to take full advantage of the efficiencies services like MaaS bring.
2.87
NRMA observed that ‘growing trends around the world point to increasing levels of ride sharing, bike sharing, carpooling, on-demand services and public transport use’:
Many vehicle manufacturers, technology companies and governments view sharing via subscription to be the most logical future for the automobile. While Australians have revered the very existence of the automobile since the early 1900s, for the first time in Australia’s history, young adults are less likely to hold a driving licence than their parents.
2.88
NRMA expected ‘private car ownership to decline as time progresses’, stating that ‘mobility will no longer be a privately-funded undertaking, but an evolving and efficient service underpinned by CAVs [connected and automated vehicles] and interconnected modes of transport’. Realising this future mobility model and its benefits required ‘integration of CAVs and traditional transport services like trains, buses, light rail and ferries’. NRMA concluded:
An automated and shared mobility future will also likely reduce congestion on Australian roads by shrinking the size of the private vehicle fleet and improving efficiency. Greater numbers of car sharing vehicles, ride sharing vehicles and on-demand taxis and shuttles will improve productivity and increase mass transport use, provided they are seamlessly integrated into existing and new services.
2.89
Transdev saw shared mobility—‘that's really what we believe in’—as ‘crucial to realising the full range of benefits of autonomous ecofriendly vehicles’. Mr David Le Breton, representing Transdev Australasia, told the Committee:
We want to avoid a future of automated single-passenger vehicles on the roads. This would just increase congestion, put more pressure on public space and reduce connection and livability in our cities. This is a pressing concern in Australia given the high level of car ownership, as we heard before; a longstanding preference for cars over public transport; and the high incidence of single-traveller journeys within relatively short distances.
Integrated transport
2.90
The key to managing the first and last mile, according to Transdev, was integrating autonomous vehicles ‘within existing public transport systems and networks—as the first mile and last mile’:
Our view is that traditional mass modes—such as rail, core bus routes and ferries—will remain the backbone of our cities' transport networks in the short to medium term. We also believe that shared autonomous vehicles will be a reality on Australia's road before single operated driverless cars. First mile, last mile autonomous shuttle services can be integrated into those existing routes to make public transport more attractive to passengers and more accessible to communities in suburban and rural areas and to community members with special mobility needs as well. As an added benefit to cities, the use of shared autonomous shuttles also reduces the demand for parking, freeing up land for more productive development.
2.91
Uber saw first and last mile solutions ‘as an important part of the future transport mix’, envisioning ‘a multi-modal transport ecosystem whereby passengers leverage point to point transport for first/last mile travel to complement their public transport journey’. Uber emphasised that ‘this integrated transport model relies heavily on strong partnerships between governments and point to point transport providers’. Uber provided examples of where it had formed partnerships with governments and other providers to provide integrated transport solutions. These included:
Partnerships with the transit authorities in Atlanta, Los Angeles and Minneapolis to provide a discount to commuters using Uber to complement public transport.
Programs such as ‘guaranteed ride home’ in Washington DC offer commuters who regularly use pooling (twice a week) reimbursement for emergency travel outside of peak hours.
In Malaysia, Grab (a regional rideshare company) partnered with the airport train service, Kuala Lumpur International Airport (KLIA) Express, to offer discounts for passengers who use Grab to reach their final destination after disembarking from KLIA Express Station in the city.
Partnerships with Mobicia, London’s leading bus times app with almost one million users per month. Uber is now integrated into the Mobicia experience, enabling customers to order a ride via the Uber app to the nearest convenient bus stop for their onward journey, improving access to public transport, especially in areas that are beyond an easy walk to the bus.
In Australia, Uber has collaborated with Transport Canberra to provide Late Night Rapid passengers with $10 discount if they used Uber to travel to and from bus stops. This was launched over the 2016 New Year period and will operate for its third year at peak times over the New Year period in 2018.
2.92
Mrs Natalie Malligan, Head of Cities, Australia and New Zealand at Uber, highlighted a recent development in Sydney, where ‘Uber was recently selected as a successful incubatee as part of the Transport for New South Wales Mobility as a Service Innovation Challenge’:
As part of this initiative, Uber, in conjunction with TfNSW [Transport for NSW], is piloting a program where riders who take an UberPool trip to or from the Manly ferry wharf within a defined geofence receive a flat $5 fare in addition to a 20 per cent discount on a connected Captain Cook ferry trip. This means riders can leave their car at home and save time trying to find a parking spot, helping reduce emissions and congestion.
2.93
Uber has also expanded its focus to new modalities. In San Francisco Uber has given people the option ‘to book a JUMP bike—an electric-assist smart bike—using the Uber app. For the first time, riders could choose seamlessly between two very different transportation modes in our app.’ Uber was also ‘working with government partners to explore JUMP bikes and micro-mobility solutions in Australia’. Uber was also exploring the development of Uber Air, ‘an initiative with the aim to create on-demand, urban aviation options via all-electric aircraft on the Uber network’. Riders ‘will push a button and get a flight via Uber Air. Uber Air will be a mass-market product serving daily and casual commuters as an alternative to driving into and out of congested urban areas’. Uber believed that ‘on-demand aviation has the potential to change the way we think about urban transportation, and radically improve urban mobility by giving people back time lost in their daily commute’.
2.94
Mrs Malligan noted that Uber was ‘focused on developing what we call Uber as a platform, our plan for an integrated future of transport where someone can push a button and get from A to B through multiple modes’:
For example, a customer journey in the Uber app could be the booking of a shared e-bike to the train station, the booking and payment of public transportation within the app, and an UberPool scheduled to pick you up at the other end. We believe integration of these public, active and shared modes of transport can offer a better journey than choosing to drive yourself.
2.95
Uber was focused on ‘asking ourselves some of the bigger picture questions about the future of transport in Australia’:
… for example, what happens if we apply innovative technology to existing transport networks? How can we extend the reach of fixed public transportation, complementing rather than cannibalising public transportation? Can tech like ours help solve the first mile, last mile problem by taking people to and from transport nodes? The short answer is that we see ourselves as part of the solution for each of these challenges, and we are also pleased to see support for transport innovations across a number of other submissions made to this inquiry.
2.96
Uber believed ‘on-demand services can also help governments provide better access to transport in a cost-effective way’.
2.97
Other groups had similar visions for integrated multi-modal transport and shared mobility. Mr Jeffries (AECOM), suggested that ‘in the centre of urban conurbations like Sydney's CBD … I think we would generally all agree that we would like to create desirable, walkable environments where people can get around on foot’. He observed that ‘in the more challenged environments, where we have a mixture of modes and a strong desire for movement and land use as well, often it comes down to very difficult and challenging decisions around how you might actually prioritise certain modes on certain streets’.
2.98
Mr Gabriel Metcalf, Chief Executive Officer of the Committee for Sydney, suggested that:
It may be that the way to transition is to do things like take a part of a city that has low rates of transit ridership, and buses that are not very well used because they don't run very often, and put in a program where you subsidise people's trips to get to the train station and you let all comers provide that service and use the subsidy—taxis, Uber, new companies that don't exist yet, shuttles. You begin to experiment with reorienting the capillary network, if you will, toward the trunk transit lines.
2.99
He observed that ‘it may be that experimenting with reorienting those patterns is more on the path than experimenting with the technology itself’.
2.100
Dr Allison Stewart, Project Director with Infrastructure Victoria, highlighted that refining the mix of services would result in different market models to meet different demands. She noted that Infrastructure Victoria had come across ‘quite a few different types of applications in our research in which people might still require private vehicles—parents with young children, for example’. She continued:
People who are visually impaired told us that they prefer to have a vehicle which they know is parked in a certain spot and where they can access it reliably. Tradespeople might need to keep their tools in vehicles. There are a lot of different scenarios in which you might need to have a privately owned vehicle, but particularly in urban areas there are a lot of reasons why you might not want to own your own vehicle.
2.101
She also highlighted the ‘move towards different types of vehicles and more flexibility in terms of how people, and particularly youth today, are thinking about the purchase of vehicles’. This approach was ‘quite different to those that have been made by more traditional generations’. She concluded:
It's quite interesting to think about how these futures might evolve with automated vehicles and with zero emissions vehicles, and how all of those different revolutions are going to change the way that we think about moving more generally—we talk about mobility solutions and we talk about that whole variety of things all happening, potentially, at the same time.
2.102
Dr Jonathan Spear (Infrastructure Victoria) emphasised that ‘different people have different preferences and needs, in terms of how they are mobile’. He observed that policy makers had ‘some choices to make about how easy we make it for people to move between different modes and to make those choices and which choices we incentivise’.
2.103
The Bus Industry Confederation issued a caution about how the mix of services was managed. A major risk of inappropriate management was social exclusion if low cost public transport was replaced by higher cost on-demand options. The BIC stated:
Roll out of MaaS and AEVs can be expected to put increased pressure on the better patronized local transit services, where demand is strongest, probably replacing them with shared car/small bus-based services, particularly when these become driverless and lower cost … This development direction reflects a blurring of the boundaries between PT as we have known it and private transport. Local transit services that have low patronage levels are at risk of losing all or most service in this context, particularly if governments rely on the market to provide most local PT-like services, expecting this to be at low cost (through MaaS with AEVs). We see this as a major risk exposure in terms of social exclusion: governments seeing MaaS/AEVs as almost the ultimate deregulation, with the market providing services to all at a very low cost. This greatly overestimates, we believe, what might be possible in terms of commercially-based service offerings in low volume markets. Risks are less if service delivery agreements are used to assure service continuity in some form, as discussed below. Fare discounts may remain for some types of passengers but there may be fewer services available locally, if patronage levels are poor, on which to take advantage of these discounts.
2.104
The BIC emphasised that ‘if service provision at the low patronage local end is left entirely to the private market place, then exclusion risks will increase, particularly in fringe urban/regional areas and in rural/regional settings, where demand densities are least supportive of commercially viable offerings for shared mobility’.
Freight
2.105
In discussions with the Committee, Mr Terry Lee-Williams, Strategic Transport Advisor with Arup, highlighted the advantages to the transport network and the urban environment of automation and alternative fuels in managing the freight task. He noted that ‘one of the great benefits of new technology is that freight and logistics can be moved in time and can be made much quieter, which means you can start to exploit the existing capacity of networks far better than we currently do’. He envisioned a future where light freight delivery and people movement could take place simultaneously in shared vehicles:
Instead of having little white vans running around, you'll just have collection centres at reasonable points, probably in underused underground carparks. You'll drop in, pick up the bunch of goods that have been deposited there by the robot to put in the back of your vehicle and, as you're going to your next customer, you drop them off. That'll lower the cost of goods but also use less space.
2.106
Heavy freight would also be revolutionised by technology:
For heavy goods, on the other hand, when you're taking pallets of food to shops and servicing the inner cities, that often rubs into the peak now. The reason for that is that they're big and noisy vehicles. I'm not a betting man, but I wouldn't mind betting that we'll get to hydrogen for heavy vehicles quicker then we'll get to battery. It's silent; there are no emissions—well, a bit of water vapour. It is feasible so long as we also take the next step and take all of those safety technologies that we're putting in vehicles right now, such as proximity sensors, automatic braking and anti-collision protection, put those in trucks and get rid of reversing beepers. Reversing beepers are another thing that'll stop deliveries occurring outside of ordinary hours because they are just so piercing and they drive people insane. You don't need them because you can't actually reverse over anything now. The vehicle will not allow you to do it. It's like an old technology that's clinging on pointlessly as we move forward. So you wipe that out and have silent vehicles. They're cheaper to operate. All the freight movement should be happening at night in populated areas. With heavy freight, say you've got a port on one side of town and a logistics centre on the other side of town. Why can't you run a B-double through the city at 2 am if it's quiet? Why not have 100 of them running through, all platooned and talking to the traffic signals? Rather than spending multiple billions on tunnels to separate these things, just separate by time, because you can be smarter about how you do those things, and then take the money you save on that and have really good passenger transport, please. Trying to squeeze the most out of every asset that we have is something we kind of forget. We always jump to the next big solution. This technology allows you to get so much more out of your existing networks.
The transition to automation
2.107
As ANCAP noted in its submission, the transition to full automation would be gradual and would involve a period of transition where vehicles with different levels of automation would share the road. DIRDC believed that operating in mixed traffic environments was inevitable and hesitated to put a timeline on achieving full automation. Mr Roland Pittar, General Manager of the Office of Future Transport Technology, DIRDC, stated:
If we look at the sensors and the processing capacity of road based vehicles, and we think of that in the context of, say, light vehicles and heavy vehicles such as buses, manufacturers are indicating to us that they feel that that sort of technology around, if I can call it conditional automation—that is not full automation in all circumstances—is probably something that will be available around the early to mid-2020s, but still requiring drivers to operate in more complex traffic environments.
2.108
Mr Pittar continued:
A business model could well be that the driver is responsible for taking the vehicle through the more complex city environment until they get to, say, a dual carriageway, and then the vehicle can undertake more of the driving task and operate in an autopilot situation. The driver might still be required by the system to take back control if there are some roadworks or if weather conditions aren't suitable, and that sort of thing. Then, at the end of the dual carriageway—the last mile, as it's often called—the driver would need to take back control to get from the dual carriageway into the next town or city at which they are dropping off or picking up passengers. That is a business model that is described to us by potential providers.
2.109
The development of automation could not be separated from the environment in which it was taking place, and policy responses needed to account for that. Ms Gayle Milnes, Executive Director, Portfolio Coordination and Research, DIRDC, advised that:
… there could be a range of different environments and a range of different policy options or tools that you could use—lane preferencing, for example, might be one tool—and you would have to think about that within a particular environment. The solution for Melbourne, for example, could be quite different to the solution for a regional centre like Wagga, for example. So it is likely that we are going to have to see how those different technologies play out or anticipate how those different technologies play out and think through what policy instruments you might use in the different types of environments. So it is not necessarily a one-size-fits-all thing.
2.110
Mr Pittar observed also that there was a need to place vehicle automation within the context of other technologies using ‘a “whole-of-system” approach’. Ms Milnes explained that the future of transport was not just about automation:
… it’s the combination or potential combination of automation of data and connectivity, of the different types of energy technologies and the shared economy. Often they have got common drivers. A common driver is the data, but it is not limited to that. Certainly the energy implications are driven by a different technology.
2.111
Mr Ian Christensen, Managing Director of iMove Australia, observed that ‘most of the technology development in that space is occurring overseas, but it is happening at all levels, from footpath delivery robots through to small trucks, through to drones, through to automated delivery, through to automated lockers, either in building basements or on the porch of your house or things like that’. He explained that there was so much diversity in the area of automation because ‘it hasn't resolved to a dominant mode of solution yet’—‘We haven't worked out yet what the optimal configuration is, nor who is going to pay for it.’ Mr Christensen noted that:
There is serious effort being made in industry to do the quantum leap, but it's difficult. In the meantime, plan B is to progressively increment the capability of vehicles as each generation of vehicles comes out, given increasing features. Both development strategies are well and truly underway.
2.112
Mr Mark Rowland, Associate, Transport and Cities Planning with Arup, highlighted one of the perils of a mixed fleet—the loss of productivity in road transport. He explained:
Some connected and autonomous vehicles that are out there now will be very conservative, so a human driver on the Monash will potentially follow another vehicle within a second. So, you imagine that brings that gap down, but what would a computer do? It would, say, 'Give them three seconds,' so then a car pulls in and then it backs off another three seconds. So, through this transition phase, you may actually end up losing capacity of your freeways when I'm talking about the mix because everybody's going to have their own risk profile. Our lanes on the Monash, for example, on a busy morning can get up to 2,400 vehicles per lane. The way to think about that is that, if every vehicle actually had a two-second gap—so, 3,600—the maximum capacity of that lane is 1,800. So, we're already driving over the 1,800. We've already got 2,400 cars an hour. So, if you add an automated vehicle in there with a three-second—and we've found this in the trials–and you back them all up, you could lose 20 to 25 per cent overnight. We've just got to be bit careful that we are going to get these amazing road safety benefits, but—this is what I mean—we're in the transition phase. We need to do trials. We need to do testing.
2.113
Mr Stephen McDonald, General Manager Strategic Initiatives with Transurban, agreed, stating that ‘a mixed fleet, in its early days, could potentially also increase congestion, as the vehicles would be more cautious in their interactions with other vehicles and that sort of thing, as opposed to how our drivers are at the moment, and I think we have to bear that in mind’. He indicated that Transurban was already looking at ‘how we prepare for some of these vehicles and how we think about our management system’. He noted that Transurban’s ‘roadways and motorways in general are well-suited to the early adoption of some of those levels [of automation] because they're nice and controlled. There aren't a lot of pedestrians, we hope, on our roads, and it's a good way to take those up.’
2.114
Dr Allison Stewart countered that Infrastructure Victoria’s research indicated that even a mixed fleet would have productivity benefits. She told the Committee that ‘in one of our scenarios, where we looked at a mixed fleet of human driven vehicles that we see on our roads today and automated vehicles of the future, we did find that overall network efficiency could potentially be improved when those two types of vehicle fleets are mixed’. She indicated that:
Even a 50 per cent penetration of those automated vehicles could result in a 70 per cent improvement in network efficiency, again, depending on how closely automated vehicles can travel to each other and how connected they are and, also, how we look at network management as a whole, in terms of where people are moving on different types of modes of transport. We expect all of those things to come together in the future, in terms of determining appropriate road space allocation on our roadways, and in terms of those connections into multimodal hubs as well.
2.115
Automation would also demand adjustments in the way drivers approached the driving task. Mr Rowland used the analogy of airline pilots to describe the adjustment that was required:
There's the automation of the airline industry. How do we learn from past examples and how would that apply in the future? We've got a lot of great examples that we can apply and learn from. Think about a pilot in the seventies and how they would have flown an aircraft. It would have been very hands-on, very direct. They would have had to interpret, but now the job of a pilot has changed. They now watch a system. They're now observing a system and how it drives on the road. Our drivers are going to have to start doing that as we move from level 1 autonomy to levels 2, 3, 4 and 5. Especially level 3 autonomy is going to be really interesting in that the job of the driver is to observe the systems; it's not to participate in the driving action.
2.116
His colleague, Mr Lee-Williams, cautioned that while the process of transition might be gradual—it might also be faster than we think:
We're still in the realm of fantasy, mostly, with people trying to prove their technologies. But remember how quickly we move with technology. An electric vehicle even three years ago was something that was for, really, high-end people with a lot of money wanting to wear a green credential. It wasn't a practical choice at all. Next year there will be practical choices available in the Australian market at price points that middle-income Australian families could easily afford. Five years from now, it'll be pretty hard to buy any mid-price vehicle that won't be electric, and internal combustion engines will be in the very low-price vehicles. So you're seeing, in a decade, a total shift in motor capacity.
Committee conclusions
2.117
The automation of transport—including mass transit—will have revolutionary implications for the way people move about. It has the potential to significantly improve mobility, especially for isolated and vulnerable people; improve accessibility to employment and services; significantly improve transport safety; and make transport networks more efficient and cost effective.
2.118
However, realising the potential benefits from transport automation will depend on the planning framework and policies that are put in place, and the vision underpinning those plans and policies. In that sense, the Committee is conscious that the recommendations of its report on the development of cities, Building Up & Moving Out, especially those relating to integrated and holistic planning, are relevant to the automation of transport. Transport automation should take place within a wider planning framework which integrates automated transport with the planning of the urban and regional environment to maximise liveability, sustainability and productivity, while acknowledging the different needs of greenfield and brownfield sites.
2.119
The Committee recommends that the Australian Government develop its strategies and plans to address the development of transport automation and alternative fuel sources through the strategic framework set out in the Committee’s report on the development of cities, Building Up & Moving Out, especially those relating to integrated and holistic planning, with a view to ensuring that transport automation takes place within a wider planning framework which integrates automated transport with the planning of the urban and regional environment to maximise liveability, sustainability and productivity, while acknowledging the different needs of greenfield and brownfield sites.
2.120
Transport automation, especially when combined with alternative fuel sources, will have significant implications for transport infrastructure. Automated vehicles will require a suitable environment in which to operate—one that they can read and communicate with. They will make demands on communications, the internet, physical infrastructure and energy networks. Meeting these demands will require careful planning and significant investment.
2.121
Optimising automation will also demand a multi-modal response—creating a seamless transport network across a variety of transport modes, connected by information exchanges (such as mobile apps) and seamless ticketing. Mobility as a Service will combine access to a variety of transport modes within a single journey, including mass transit, shuttles and active transport. To achieve this, we must rethink how public transport is organised and contracted. The concept of a systems approach, in which operators provide services across areas rather than modes, with shared mobility contracts and app based booking, could transform urban transport. However, this will require coordination with private transport and rideshare operators and a commitment to social inclusion.
2.122
Within this context, mass transit has an important role to play providing high-volume trunk routes as the core of the transport network. There are significant opportunities for the use of automation in the provision of rail, light rail and buses services. The automation of rail is already well advanced and the development of autonomous trains should be continued. The automation of buses is well advanced as a technology, but its development as an operating system in a real-world environment is still in its infancy. The possibility of using automated buses in dedicated busways should be explored as a key first step to wider automation. The Committee notes that governments are already involved in the trial of shuttle buses to improve mobility and first and last mile transport situations, and supports this work being advanced.
2.123
Achieving the benefits of automation will require explicitly targeted policy responses. If the ideal is compact and accessible urban environments and well connected regional areas, then policies around transport automation need to explicitly support these outcomes. The goal should be the creation of a new transport ecosystem. Consideration should be given to policies which promote the development of this ecosystem, including restrictions on private automated transport and ridesharing, and the encouragement of shared mobility based on strong trunk routes provided by rail, light rail and buses, connected to smaller vehicles providing connectivity within cities and suburbs. There also needs to be a deliberate effort to incorporate active transport into the network. Achieving coordination between different organisations and modes is essential.
2.124
The Committee recommends that the Australian Government adopt as its goal support for the development of a new automated transport ecosystem, incorporating shared mobility based on: strong trunk routes provided by rail, light rail and buses, connected to smaller vehicles providing connectivity within cities and suburbs; active transport solutions; and strategies to manage the use of private automated transport and ridesharing in city centres and heavily congested areas and routes.
2.125
Further recommendations regarding these issues are presented in Chapter 4.
2.126
While this has not been a central theme of the inquiry, the Committee is also aware of the potential for automation and electrification to substantially improve the management and handling of freight—light freight becoming integrated with shared mobility passenger movement and heavy freight transport becoming quieter and more efficient thanks to electric motors and platooning of vehicles. Alongside passenger transport, the benefits of automation and electrification of freight transport should be considered by governments.
2.127
The Committee recommends that the Office of Future Transport Technology within the Department of Infrastructure, Regional Development and Cities undertake consideration of the benefits of automation and electrification for the transport of freight.
2.128
The art of transition will be a key factor in the success of automation. The transition to automation will involve mixed fleets of vehicles of varying degrees of automation travelling together. Managing this transition will require regular assessment and adjustment of the regulatory environment surrounding road transport, vehicle specifications (e.g. safety distances between vehicles) and driver behaviour. Public understanding of the changing road environment is essential to a safe and successful transition to automation.
2.129
The Committee recommends that the Australian Government, in conjunction with State and Territory Governments, develop a strategy for managing the transition to full automation on roads, including mapping regulatory responses, vehicle specifications and driver training requirements.