PRT’s CONUNDRUM

Why the business case for personal rapid transit is blocked.

Lawrence J. Fabian, Trans.21 and Stan Young, NREL

May 2016

Urban transport policy-makers have new street-based mobility options. The emergence of robo-cars, robo-cabs, robo-delivery vehicles, robo-vans and robo-buses is upon us. How real and safe and affordable they will be remains to be seen. The major economic appeal of such possibilities -- accompanied by ride-sharing and community-based mobility networks -- is that they use already existing streets.

The Morgantown PRT has served
West Virginia Universitysince the 1970s.

Driverless cars are bringing fresh air to discussions of envisioning and supplying urban mobility. They clash with PRT promoters who speak with bravado and hypothesis. They promise a sort of mobility paradise, often with references to far-off places like India, China and Sweden. Upon closer inspection, these are often little more than carrots for investors.

PRT promoters speak in the conjectural tones of “could” and “would”. Or “can” and “will”.   After 50 years of such conversations, we cannot truthfully state in the present tense: PRT satisfies urban transport needs. At best, we can point to West Virginia University, which is served by a USDOT demo from the 1970s and a few recent shuttles overseas.

The Project Case Dilemma

Then and today the conundrum of PRT is in the very business case that must be made. Only a large network brings out the potentially significant benefits of non-stop service scheduled in real time that is possible with off-line stations.

By Merriam-Webster, a conundrum[1] is a confusing or difficult problem that has only a conjectural answer. What blocks PRT interest is that a large commitment to extensive guideway and stations is necessary long before the benefits are felt. As common with large infrastructure projects, big bucks go out long before substantial payoff.

An early plan for Fleminsberg, Sweden had
about twenty stations.

Urban residents, elected officials and local technocrats know that most US cities abound in streets, alleys, utility rights-of-way, and local arterials, much of them hardly used. In countless residential blocks and old retail and industrial districts that predate Eisenhower’s Interstate Highway System which reoriented urban development around interchanges, there is excess street capacity.

Adjoining many stretches of multi-lane Interstate and other expressways are little-used local streets. They offer ample opportunities for local robo-services. In contrast to PRT’s conundrum, the scenario of robo-cars shaped for public good becomes compelling, especially where icy winters are not a factor. An entrepreneur can start without major investment in infrastructure. Expansion - perhaps with guideway for longer trips -- comes later if and when desirable.
  

Ordering Our Thinking

NHTSA in 2013 published an official classification of levels of vehicle level of automation. It is well used in technical transport discussions. Yet we lack a comparable classification of street environments within which specific robo-services are to operate. The service domain can be in private property or civic campuses. ITS and AUVSI engineers increasingly talk not just of vehicle-to-vehicle communications. They foresee two-way communications between vehicles and fixed infrastructure that is itself intelligent and interconnected, with data flowing to a control center somewhere creating “big data”.

Someday a national center in Omaha or in the “cloud” may super-monitor all street movements. That is a way far off in the future. For now, it is helpful to establish a policy-friendly categorization of the two-dimensional, geographic areas or domains over which specific robocar services will operate.

Managers of private campuses of a square mile or two, for example, have pretty full control over their own property. A fleet of robo-cars that go no faster than 15 or 25 mph will do just fine.  They don’t need to put them out on the busy highways. The fleet can remain on campus roads, paths and parking lots. This is Level 0 of robo-domains: limited and controlled, private property (including some governmental owners like ports and the military).

Enter the Smart Context

Remarkable progress is being made in sensing and visioning systems. Cutting-edge thinkers are developing visions of smart cities with electronic markers dotting the landscape to facilitate (or duplicate) vehicle tracking and controls.

Smart vehicles watch out for cars, kids and other obstacles on the running surfaces around them. Smart infrastructure can tell them a lot more. Local markers can communicate real-time network conditions upstream perhaps diverting vehicle to a better itineraries. Simple smart infrastructure can be confined to a parking complex, or serve fleets circulating within large university and medical complexes. More ambitious thinking takes on large-scale traffic management for entire metro regions.

USDOT’s SmartCity Challenge aims to bring these potentialities to a single demo city.

Two-Dimensional “Robocaricity” Projects

Local mobility companies can
operate at Levels 1 or 2, here
in cooperation with the transit agency.

The performance and safety of robocars in cities will increasingly depend on smart infrastructure. There is a significant scaling effect. How large and complex are the service domain?

Here are six categories of Robo-Service domains:

                Level 0   Secured, managed private or semi-private campuses

                Level 1   Supervised activity centers, probably mixed uses that rely in part on public streets

                Level 2   Urban districts with firm boundaries (physical or by policy) with residential and/or employment populations of, say, 5000 to 50,000

                Level 3   An entire city - central or suburban - but within the jurisdiction of a single legal entity or municipality, aka City Hall

                Level 4   A metropolitan area with populations of, say, 0.1-10 million

                Level 5   A very large metropolitan or inter-regional with significant long-distance trips (say, 10+ miles). In a few decades, policy makers may want to think about a sea-to-sea domain as the 6^th and international as the 7^th levels.

---

[1] The origin of this world seems to be mockery of intellectuals who mixed Latin into their 16^th century English.