Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
The writer is the founder of Khosla Ventures
Imagine a world that operates by public transport in which your ride arrives the moment you need it. You travel directly to your destination — no stops, no transfers — at a fraction of the cost and environmental impact of both individual cars and high-speed trains.
Best of all, this kind of system doesn’t require us to tear up our streets. It fits within the margins of our existing infrastructure, and would only require marginal incremental infrastructure investment.
Self-driving, on-demand robotaxi pods could change the way we move around cities and dramatically increase throughput of our congested streets. We should be building fleets of these lightweight, low-cost electric vehicles.
Known as personal rapid transit systems (PRT), they can travel along narrow, dedicated pathways closed off to all other forms of transport and passengers can get on and off them anywhere they choose.
Earlier this month, Elon Musk’s Tesla unveiled its plan to create autonomous robotaxis. Alphabet-owned Waymo already operates robotaxis in some US cities. But this vision is not ambitious enough. It will further congest streets, just as ride-hailing company Uber has done. It accepts the existing limitations of our streets without marrying it to a network of dedicated lanes.
This is because many people still find it hard to imagine life without cars. In the mid-20th century, the US made a pivotal choice that shaped its cities, economy and lifestyle. It chose highways and cars over public transit. At the time, this seemed like the future: the freedom of the open road, the allure of suburban living, and the booming post-second world war economy all converged to push America towards a car-centric culture.
The Federal-Aid Highway Act of 1956 cemented this vision, unleashing a highway system that encouraged suburban sprawl, fuelled the automotive industry and sidelined public transit. Rail systems were seen as relics of a slow, industrial-era technology ill-suited to America’s postwar aspirations. The car was king.
But this congested system is breaking. In 1950 about 30 per cent of the world’s population lived in cities. By 2030 this is expected to reach 60 per cent. Infrastructure cannot keep up with this growth. An increase in cars further reduces street capacity.
We need new solutions. Public transport systems like trains (light and heavy rail) are expensive to expand and inefficient when they are not full. Fifteen years ago, I ran the numbers and found that electric rail systems in the Northeastern US had worse carbon emissions than gasoline-powered cars. When applied to an autonomous, ride-sharing model, the difference in per-passenger energy use is even greater. This is largely due to average load factors. Outside peak times a passenger train occupancy hovers around 30-40 per cent.
The energy use of high speed electric trains in the US is 200-600 watt-hours per passenger mile (Wh/PM) over a 24-hour period in an urban environment. For an autonomous EV pod carrying four passengers it might be around 50 Wh/PM.
There is also the question of disruption. The streets in our cities are already as wide as they’re going to get. We cannot expand roads for new transportation solutions without incurring massive costs and disturbance.
PRT pods can operate on comparatively tiny — bike-lane width — dedicated pathways. The cost per passenger mile is significantly lower than for individual cars and high-speed rail, for both the vehicles and the infrastructure. While high-speed rail requires extensive, expensive tracks, signalling systems and station infrastructure, these systems can be deployed on simple networks, often using existing roads or new, lightweight guideways that are cheaper to build and maintain. Eventually little public funding will be needed.
A version of the pods already exist at some airports. Startups like Glydways, of which I am an investor, is winning contracts and hopes to bring the systems into use in areas like San Francisco’s East Bay.
But there needs to be more. There are around 200 cities in the world with metro systems and thousands of cities large enough to require them. The problem is that they cannot afford the existing options. PRT changes the calculations. It is but a fraction of the cost of light rail and bus to build and operate.
Infrastructure projects don’t need to get bigger to be better. Autonomous, public robotaxis provide the best of both worlds: smaller, more flexible vehicles that adapt to demand in real time, with on-demand point-to-point service, and infrastructure that doesn’t break the bank.
