The US Department of Transportation is presently soliciting input on the above topic. Our response to some of their subtopics is presented here.
1. Role of government
Federal government should work to understand how market forces and regulatory requirements may unbalance the playing field to the detriment of society. Some examples follow, comparing autonomous vehicles to personal rapid transit (PRT) where PRT is a systematized form of autonomous vehicles travelling on separated fixed guideways and functioning as a public transit system, while autonomous vehicles function individually and travel on the open road.
If autonomous vehicles can increase road safety and capacity, significant benefits will accrue to the travelling public. PRT is already proven to reduce road congestion and to be orders of magnitude safer than cars and transit and, if it also proves to be scalable, significant benefits will accrue to the traveling public. It seems that both technologies (variations on the same theme) have the potential to significantly benefit the travelling public and will probably each address different segments of the market that may blend over time.
a) Market forces
Since autonomous vehicles will mostly be sold to the general public and since they are being developed as incremental modifications of conventional automobiles, the barriers to entry are not high. In addition, since most households own cars, the general interest in autonomous cars is naturally high and many already dream of owning one.
On the other hand PRT is a system that requires new infrastructure. Even though the total cost of a PRT system per passenger mile can be much less than the total cost of the road/automobile system, this requirement for new infrastructure is a huge barrier to entry. It means that most PRT systems will be purchased by government agencies and require significant upfront capital funding. Federal government agencies tend to react to requests from their constituents, many of whom have never heard of PRT and do not understand its potential to solve their problems. In addition, the political situation is such that a mayor would often rather wait for a $1 B light rail project with 50% FTA funding than proceed immediately with a $400 M locally- funded PRT project that does a better job but does not bring $500 M of federally funded jobs to his/her community. Even if the PRT project were FTA funded it would only bring $200 M federal money to the community. The current federal subsidy of fixed guideway projects is thus a barrier to entry for new, innovative and less expensive solutions.
In addition, federal funding of transit projects is biased towards conventional corridor-based modes since many qualification criteria are based on corridor analyses. PRT can function in a corridor but, in many applications, provides optimal service when functioning as a network. Since most people desire to travel from origin to destination with little or no need to pass through a corridor on the way, it would seem more appropriate to consider transportation alternatives on the basis of service area analyses rather than corridor analyses.
While it is appreciated that the scope of this RFI has been broadened to include PRT, this is insufficient. The scope needs to also include the driverless transportation of goods and freight – preferably not limited to roads but also including railroads. The driverless revolution will impact all of these modes and the interaction between modes must be understood if full advantage is to be taken of automation. The ability for automation to allow us to reinvent the entire surface transportation system (and some aspects of the air transportation system too) should not be underestimated.
Automating cars in mixed traffic is a challenge. However, the mix of small and large vehicles may change dramatically if drivers are eliminated. To what extent will it be necessary to have large vehicles transport freight if all vehicles are driverless? What proportion of freight would be more economically and efficiently moved in small driverless vehicles that require much smaller and less expensive supporting infrastructure? How much less capital and O&M cost would be required for the infrastructure? What would the impacts on road congestion be?
The role of the federal government should be to understand the societal benefits of the different solutions and seek to balance market forces in a way that society receives the solutions it needs, not just the ones that natural market forces and political considerations tend to favor. The federal government needs to lead in the development of an integrated automated surface transportation system for people and freight that dramatically improves safety, efficiency and sustainability. Such a system needs to address all of the following issues related to surface transportation:
• Energy use
• Foreign oil use
• Emissions and climate change (cradle to grave)
• Overall mobility/accessibility
• Mobility/accessibility of non-drivers
• Mobility/accessibility of non-vehicle owners
• Severance of neighborhoods
• Right-of-way requirements
• Capital, operating and maintenance costs
• Economic impacts
• Land use
Specifically with regard to PRT, one role of the federal government should be to fund a PRT demonstration program. While many of the concerns expressed in the recent San José PRT feasibility study may be easily addressed, some, such as scalability to large networks, require complicated hardware-in-the-loop evaluation, and almost all could be overcome with an extensive demonstration program. Without such a program, PRT in the U.S. will continue to languish behind systems being deployed elsewhere and will likely be limited to small campus-type applications for many years. A PRT demonstration program is needed in order to demonstrate if PRT is scalable to widespread urban deployment of the kind necessary to attract many drivers from their cars – a feature that could probably eliminate congestion long before autonomous vehicles are able to do so. If PRT can be widely deployed, its ability to significantly increase transit mode share has been demonstrated in many studies such as by Dekhordi which show transit mode shares doubling, tripling and more in many European cities.
b) Regulatory Requirements
Cars have a horrendous safety record yet the general public seems to have grown immune to the dangers of driving – they are presently outraged at gun-related deaths but seem not to realize or care that even more are killed by cars every year. On the other hand the Morgantown PRT system has completed over 140 million injury free passenger miles in its 37-year existence – a safety record unmatched by any non-automated transportation system.
The regulatory requirements for PRT in the U.S. seem likely to be based on the ASCE Automated People Mover Standards which seems reasonable since PRT is a subset of automated people movers. The safety requirements in these standards are very rigorous and will result in PRT systems far exceeding the safety of driven systems.
The regulatory requirements for privately-used autonomous vehicles seem unclear but may be no stricter than those for present-day automobiles. Autonomous taxis on the other hand could be regulated as automated transit just like PRT. This could create an interesting situation where privately-used autonomous vehicles come to fruition quickly, but do not bring substantial safety benefits and autonomous taxis struggle for a long time to meet the rigorous standards PRT currently meets.
The role of the federal government should be to evaluate the benefits and costs of regulatory requirements w.r.t. safety (and other issues), to determine appropriate requirements and to ensure that all transportation systems provide similar levels of safety. The current disparity between automobile and airline safety should be eliminated (and not by decreasing airline safety!).
2. Smooth assimilation of automation into an integrated transportation system
PRT is leading the way in automating passenger vehicles because it is doing so in a controlled environment – automating vehicles on the open road is far more difficult and being undertaken in an uncertain regulatory environment. Autonomous vehicle developers could learn from the experience of the PRT community.
Smooth assimilation of automation into an integrated transportation system will not occur without a rigorous systems engineering process. This process must start at the highest level and incorporate all surface transportation. The federal government needs to initiate a process whereby the requirements for an integrated fully-automated transportation system of people and goods are developed and agreed upon. Only once we understand the end system that we are striving for can we know how each component should be developed and can we design a logical phased implementation plan. The opportunity exists for the U.S. to develop a revolutionary automated surface transportation system that is highly efficient and safe. It will probably require a Kennedy-like vision and sense of mission to accomplish this.