It may not have been the most thrilling getaway job, but San Francisco law enforcement said it’s one of the first crimes of its kind. It also remains unsolved after nearly six months. According to officials speaking with the San Francisco Chronicle, police are still investigating a case in which an unidentified man stole an arm-full of activewear from a local yoga studio, then fled the scene inside a self-driving Waymo taxi.

“I would think it would be easier to solve in a Waymo,” Sgt. Tim Faye told the newspaper on June 4.

Anticipating an open-and-shut investigation is understandable, but the situation is actually more complicated than it seems. While police couldn’t comment on an active case, it’s almost certain the robber used a burner account or stolen phone to order the taxi service, which debuted its self-driving option to San Francisco customers in June 2024. The white Jaguar used during the getaway features around 29 high-definition cameras mounted inside and outside the autonomous vehicle that provide 360-degree vantages, but Waymo only temporarily retains recordings. The company erased all interior footage by the time investigators filed a search warrant in April 2026, and data privacy laws ensure that any faces captured on cameras outside the car must remain blurred.

Skeptical customers and safety concerns have restricted autonomous ridesharing to only seven cities across California, Arizona, Texas, Florida, and Georgia so far. That said, companies like Waymo are still aggressively pursuing plans to expand the service to other parts of the country. One of the only other similar crimes occurred last year in Los Angeles, when a suspect allegedly robbed a grocery store and then left in a Waymo. In that instance, police pursued the vehicle and successfully pulled the car over after turning on its emergency lights.

“It’s highly unusual in the first place that a Waymo is even used by a suspect,” added Sgt. Faye.

The post Police can’t find shoplifter who fled in self-driving Waymo appeared first on Popular Science.

In 1953, Donn Fichter, a graduate student at Northwestern University in Chicago, had a simple transportation idea: What if you tipped an elevator on its side, enabling it to run horizontally, and set it loose in a city? Unlike conventional urban mass transit, elevators are responsive to individuals, callable with the push of a button, and not subject to schedules. 

After completing his dissertation in 1958, “Automated Urban Circulation,” Fichter spent years turning that idea into a complete transit system design he called Veyar. At its core, Veyar would offer small automated cars running on lightweight guideways. The electric cars would be available at any hour and travel directly from origin to destination without stops, schedules, or drivers. “Personalized transit,” he called it, in which each car “is a self-operating vehicle which can go unattended.” To keep infrastructure construction costs low, he explained, “they have to utilize existing public right of way: the streets.” Fichter self-published the design in 1964, calling it “Individualized Automatic Transit and the City.”

For 60 years, personalized transit systems like Veyar gained support from generation after generation of transportation engineers, but none were ever built. That’s because personal rapid transit systems demanded infrastructure cities couldn’t afford and automation technology that didn’t yet exist. What finally solved both problems wasn’t a transit agency or a federal program, but rather the autonomous vehicle industry. Companies like Zoox and Waymo built Fichter’s system more practically, starting with the automation and letting existing streets serve as the guideways.

The origins of personal rapid transit

Donn Fichter was born in Minneapolis in 1926. After serving in the Army during World War II, he earned engineering degrees from Brown and Northwestern. He was the first serious advocate of what urban planners would eventually call personal rapid transit, or PRT—a vision of on-demand, automated, point-to-point city travel. 

Fichter conceived Veyar at a time when traffic choked American cities. Cars gave riders individual freedom at the expense of gridlock. Buses, subways, and elevated rails offered more efficiency but subjected riders to fixed schedules and routes. 

What no one had yet built, Fichter argued, was a third system that combined the automobile’s spontaneity and the subway’s separation from traffic, available to anyone at any hour without a driver, a schedule, or a transfer. Gridlock notwithstanding, the environmental stakes, he believed, made the solution urgent. 

Even before the first Earth Day was celebrated in 1970, Fichter foresaw the “ecological imperative” to reduce our dependence on private automobiles. He made this case explicitly in a 1968 PRT planning paper, “Small Car Automatic Transit.” Fichter claimed that small, electrically-propelled, autonomous cars running on guideways would mean cleaner air, quieter streets, and cities less congested with the machinery of driving and parking.

Personal rapid transit systems catch on in the 1970s

“Your car is waiting,” wrote journalist Paul Wahl in a 1971 Popular Science feature on personal rapid transit systems. “On entering the car, you push a button to select your destination, then take a seat. The cabin is roomy, automobile-like in accommodations.” 

Wahl went on to describe how a PRT trip might unfold: “The automatic vehicle moves away on the station spur, accelerating until it enters the stream of traffic on the guideway.” The car would then switch off the guideway at its destination spur, “with a central computer doing the driving.” 

The experience Wahl described was precisely what Fichter’s Veyar system proposed. Small electric cars—sized for just a few riders—would run on slender elevated tramways threaded along existing streets. Stations every few blocks would keep cars queued and ready. Just like an elevator, a rider would board, close the door, press a button, and go. The car would merge into mainline traffic automatically, travel nonstop to the destination, and pull itself into the arrival station without further instruction. Then it would wait, callable for whoever needed it next. A computer would control the entire network. What the elevator had done for the skyscraper, Veyar would do for the city.

The federal bet on personal rapid transit begins with the Nixon administration

By the early 1970s, the idea attracted serious attention. As Wahl wrote, experts were “banking on it to relieve our metropolitan areas from the twin stranglehold of pollution and congestion.” The federal government committed $6 million to build and demonstrate four competing PRT systems at Transpo72, an international transportation exposition held at Washington, D.C.’s Dulles International Airport in 1972. One of those prototypes was destined for a small college town in West Virginia, where West Virginia University needed a better way to move students between its multiple campuses and downtown Morgantown, West Virginia. 

At the same time, planners in Minnesota began drawing up blueprints for a city to be built from scratch on 50,000 acres of rural land—a place called the Minnesota Experimental City, or MXC. The new city was the brainchild of Athelstan Spilhaus, a polymath University of Minnesota dean who had already helped design the 1962 Seattle World’s Fair and co-invented submarine warfare instruments. Spilhaus wanted MXC to be a living laboratory, not a utopia, and personal rapid transit was to be its arteries. 

The federal commitment to PRT in the early 1970s produced a brief but remarkable flurry of competing designs. Engineers at aerospace firms, university labs, and automotive companies developed more than two dozen distinct guideway systems—Monocab, TTI, Dashaveyor, Cabinentaxi, Aramis, staRRcar, and others—each with its own switching designs, propulsion method, and structural approach. No two were compatible. The proliferation reflected a significant engineering problem—no one had cracked the code on the automated control systems required to make PRT work. Wahl called this control system the “super-robot trainmaster.”

“The heart of any personal rapid transit system,” Wahl wrote, “is the central computer facility that runs things efficiently and economically, making it practical.”  He described a fully autonomous system that not only controls all the cars, “but also handles vehicle distribution and scheduling.” In fact, the central computer would manage just about everything, he explained, leaving little to human operators who are prone to make mistakes. Unfortunately, at the time, such sophisticated automation technology did not exist. 

Besides lacking the necessary automation, PRT systems demanded infrastructure cities couldn’t afford to build at scale, even with available federal funding. A network of lightweight guideways would need to be built above city streets, with stations every few blocks, for PRT to deliver on its promise. By the mid-1970s, federal funding had dried up, Transpo72 had come and gone without producing a single municipal contract, the Minnesota Experimental City project had been canceled, and PRT’s moment of official enthusiasm had passed—with one notable exception.

America’s first and only personal rapid transit system

The West Virginia University Personal Rapid Transit system, which opened in Morgantown in 1975, became the closest thing to a guideway-based automated transit system ever built for regular urban use in the United States. It connects the university’s three campuses and the downtown central business district via 8.7 miles of dedicated guideway and five stations, carrying riders in small electric vehicles on demand, without stops between origin and destination. And most importantly: The system works. 

Since its debut in 1975 WVU’s personal rapid transit system has logged more than 100 million trips, using electric vehicles that carry roughly 12,000 passengers a day during the school year. Despite its impressive track record, Morgantown also illustrated the trap at the heart of every PRT proposal. The project ran wildly over budget—partly because engineers rushed to meet a politically mandated deadline tied to the Nixon administration—and the cost per rider was never remotely competitive with conventional mass transit. 

More fundamentally, Morgantown succeeded because it was built for a specific, constrained geography: a university town with four fixed nodes and a captive ridership. That configuration bears almost no resemblance to the open-city, go-anywhere network with stops every few blocks that Fichter had imagined, and it offers no blueprint for replication in a traditional urban setting. For a major city to build what Fichter described, it would have had to retrofit onto automobile-centric city streets dozens or even hundreds of miles of elevated guideway. It’s something no city has ever tried.

Driverless cars: PRT without the tracks?

And yet, six decades after Donn Fichter sketched his first Veyar pods, you can summon one of their descendants with your phone. Today, Waymo operates driverless electric vehicles across six major American cities, completing nearly half a million rides per week in 2025. 

Amazon’s Zoox has deployed a uniquely designed robotaxi—no steering wheel, no pedals, carriage seating for four, bidirectional so it never needs to turn around—on the streets of San Francisco and Las Vegas. Between them and a growing field of competitors, the age of individualized automated transit has arrived—just not as anyone planned.

But do robotaxis really fit Fichter’s vision? A car can be summoned with the push of a button. It travels straight from origin to destination without stops. It is “a self-operating vehicle which can go unattended” as Fichter described Veyar in 1964. 

Fichter would recognize robotaxis instantly as personalized transit. What’s missing is the “rapid” promise of a PRT system. Driverless taxis are subject to the same traffic-choked congestion that has plagued American cities for nearly a century. 

In his 1964 specifications, Fichter worried that driverless vehicles “could not expect to share the streets with other motor vehicles,” which is why he proposed elevated guideways. Today, his concern seems prescient. 

Waymo has faced recalls for vehicles driving into flooded roadways, investigations into repeated failures to yield to school buses, incidents where robotaxis blocked emergency responders at active crime scenes, and acted as getaway cars. A citywide power outage in San Francisco in 2025 triggered a wave of vehicles simultaneously requesting remote confirmation checks, snarling traffic for hours. The riding experience remains geofenced to specific neighborhoods in specific cities. 

When issues arise, the system relies on remote human operators—Waymo employs about 70, half of them based in the Philippines—to step in. But these are engineering problems being worked through, not evidence the concept is broken. Arguably, city streets become the guideways when they are filled almost exclusively with robocars, which would complete Fichter’s vision in spirit, if not in intent.

But robotaxis were built as a for-profit product, not as civic infrastructure. They are privately owned, unevenly distributed in cities, expensive on a per-ride basis, and poorly regulated across most of the United States. 

What Fichter envisioned was a public system woven into the city—the way elevators are woven into buildings—affordable to everyone, and available at the push of a button. Waymo, Zoox, and their competitors have built something remarkable. But whether it someday resembles the civic infrastructure Fichter had in mind, or remains just another profit-based enterprise siphoning riders and revenue from transit agencies, is ultimately a policy question—one that cities and regulators have so far shown little urgency to answer.

In A Century in Motion, Popular Science revisits fascinating transportation stories from our archives, from hybrid cars to moving sidewalks, and explores how these inventions are re-emerging today in surprising ways.

The post Robotaxis almost happened in 1964—with help from the U.S. government appeared first on Popular Science.

When most American travelers conjure up the image of a bus, many words come to mind but fast almost certainly isn’t one of them. An ambitious proposal in California wants to change that by exploring the idea of buses operating between 100-140 mph.

Though buses function as an integral means of affordable transit for millions of people, they certainly aren’t the mode of travel for anyone in a hurry. Long-distance operators like Greyhound, traveling from city to city, typically max out at 65 miles per hour, and frequent stops mean a bumpy trip aboard one can easily take twice as long as the same journey by car.

But what if that same bus could reach speeds rivaling a train? That’s an idea currently under consideration by California’s Department of Transportation (Caltrans), which recently held a webinar discussing the feasibility of a “freeway bus service”—a concept envisioning a new fleet of specialized buses traveling down an interstate at speeds approaching 140 miles per hour. These so-called “bullet buses” would have their own dedicated high-speed lanes and could theoretically transport dozens of passengers from San Francisco to Los Angeles in around three hours and connect smaller rural communities along the way. That same trip on a long-haul bus today takes somewhere between seven and nine hours.

In an email to Popular Science, Caltrans emphasized that its interest in the buses remains very much in the exploratory phase. And while building out such a system would require significant time and financial investment, the agency describes it as “conceptually feasible.” But are these speedy buses actually a good idea and would they even work? In a vast graveyard of failed public transit proposals, could bullet buses buck the trend? And even if they are built, would anyone want to ride in one?

Bullet buses would require sleek new vehicles and wider roads 

Caltrans detailed the prospect of an interstate high-speed bus system in an 18-page report released last year. In it, they envision a dedicated high-speed bus lane connecting cities and rural areas. This multi-purpose lane could serve local routes (stopping every two to four miles), express routes (stopping only at interchanges), or long-distance routes (traveling between cities). The concept essentially applies the tiered service model already used in subway and rail systems to long-distance buses. In theory, this additional transit option would simultaneously put fewer cars on highways and reduce the burden on the state’s still-developing high-speed rail network.

That’s the hope, at least. Getting there would require fundamental changes to both the design of buses and freeway infrastructure. Current freeways are typically only engineered to support speeds of only up to 85 miles per hour. And while plenty of speed demons exceed those limits daily, the caps aren’t just theoretical; they directly shape design considerations like a road’s curve radius and camber, the slight banking that helps vehicles stay stable through turns.

In other words, vehicles traveling at 140 mph on current roads would have far less ability to safely navigate what lies ahead and would struggle to maintain control. Those risks, the report rather drably acknowledges, would make any collision “catastrophic” at those speeds “given the low survivability.” Seatbelt use, they add, would be mandatory. 

All of this means a high-speed freeway bus service would likely require roads redesigned from the ground up. The bus lanes alone, the report notes, would need to be at least 12 feet wide, with an additional 12 feet for both the inside and outside highway shoulders. Entry and exit ramps would also need to be significantly longer to accommodate the higher speeds.

Then there’s the actual buses. Simply adding a turbocharger to the currently in-service bus won’t cut it. These bullet buses would similarly require entirely new design. This updated approach would have to consider factors like drag, turbulence, and airflow, with the exterior chassis constructed in a shape far more aerodynamically efficient than the boxy brutes on the road today. Ironically, these sleeker more curved designs would likely look similar to bullet trains. Carbon fiber would also probably feature prominently to reduce overall weight. 

But the bus body isn’t the only thing that would need an overhaul. The report notes that the brake systems currently used wouldn’t work. Those brakes generally perform well at speeds up to 88 miles per hour, but failure rates rise substantially above 90 mph. And bullet buses would travel considerably faster than that. They would also need new specialized tires capable of withstanding the additional heat and stress that are primary drivers of blowouts.

Bus operators would also look different. Though a human would likely still be needed to board passengers and handle local road driving, the report suggests that travel on high-speed lanes would be handled autonomously. Human drivers, the argument goes, simply don’t have fast enough reaction times to safely operate these bullet buses. And while self-driving vehicles are an increasingly common sight on US roads (and even some Texas highways) none of those systems currently operate anywhere close to the speeds bullet buses would need to reach.

“Pushing a bus to 100–140 mph requires a re-engineering of the vehicle: high-speed rated tires, extremely powerful brakes, active suspension and stability control, aerodynamic streamlining, lightweight but strong construction, and robust safety systems,” the Caltrans report notes. 

Turning all that into reality, especially in an environment where new mass transit efforts notoriously face backlogs and delays, will be challenging. But some experts see real potential upside. DePaul University Professor Joseph Schwieterman, an expert in transportation and urban planning, told Popular Science these fast trains could potentially fill in certain gaps where high speed rail falls short. Buses, operating on roads with wheels, can intimately handle sharper turns easier than trains operating on a fixed track. Buses, even those operating at high speeds, can also accelerate and decelerate much easier and faster than rail. Those factors combined with the ability to operate on already existing road means the buses could potentially get far more passengers closer to to their destinations than spread apart rail stations 

“The concept is intriguing because fast-running buses could complement high-speed rail service, so it is not an “either/or” proposition,” Schwieterman said. “Fast buses are likely to eventually be part of the mobility ecosystem. But the lack of real-world examples of high-speed buses in operation makes California’s high-profile discussion about the technology seem premature.” 

Schwieterman also expressed some skepticism over whether or not the average traveller would necessarily embrace the idea of strapping into an ultra fast bus with open arms. Those with a possibility toward car sickness may also view these travel methods as something out of a nightmare. 

“The evidence is clear that many intercity travelers are reticent to travel by bus on trips longer than three hours,” he said. “The interiors of buses could be configured to support first-class service, but there would still be much uncertainty about the traveler response.”

“The effects of swaying over curves could be particularly troublesome,” he added 

International attempts at faster buses  

The fast train network proposed in California draws some inspiration from a handful of international alternative bus systems, but none have come close to hitting 140 mph over prolonged periods of time. 

In the 1970s, the Brazilian city of Curitiba built what it calls Bus Rapid Transit (BRT), a system that uses dedicated bus lanes to transport large groups of passengers long distances. Though these buses only ever approach a maximum of around 60 miles per hour, the dedicated lane means they function similarly to an above-ground subway line. Today, more than 2.5 million passengers across 200 cities use it daily.

Around that same time, halfway across the planet, engineers in Adelaide, South Australia, constructed the “O-Bahn,” a series of high-speed, guided buses that run on tracks. The unusual design essentially takes a standard commuter bus and plops it atop dedicated concrete rails normally intended for trains. This hybrid approach means the buses can cover long stretches quickly without any traffic, and then leave the tracks and use standard wheels to drive on regular roads for more local routes.

In terms of pure speed though, the closest example comes close to what Caltrans envisions in the “Superbus” from the Netherlands. This one-off prototype looks like a cross between a bus, a Formula One car, and a rocket ship. Its speed reflected that. In tests, the all-electric, 23-person black tube could reportedly reach 155 miles per hour. Its sleek, racing-inspired design had 16 gull-wing doors and a drag coefficient similar to that of a super car. But even though the Superbus proved it was at least conceptually possible to move bus quantities of people at high speeds, the project fizzled out because it would have required entirely new “super lane” roadways to be built out for it. Today, the lone superbus gathers dust in a University of Delft warehouse. 

The bullet bus has a bumpy road toward reality 

It’s still unclear how far California wants to pursue its high-speed bus vision. In an email, Caltrans told Popular Science it’s currently “evaluating what would be required before determining whether future testing or implementation” is appropriate. But technical feasibility is only part of the battle. Getting public support for the bus system would also likely face an uphill battle, especially since the state’s now decades-long plan to build a high-speed rail network connecting Los Angeles to San Francisco is still nowhere near completion. That route, once expected to cost $33 billion and be finished by 2020, now has a price tag exceeding $100 billion for a substantially shorter route.

Schwieterman, though optimistic about the concept of a high-speed bus network, said engineers need to slow down and iron out many more specifics before plowing forward. 

“I believe the idea should be quietly dropped until California or another state tests the workability of fast buses in a controlled environment,” Schwieterman said. “Starting with, say, a 50-mile route where buses reach 100 mph and ramping up from there would be more practical than engaging the public now in a debate about ultra-fast buses on long-distance routes.” 

The post Forget high-speed rail: California is exploring 140 mph bullet buses appeared first on Popular Science.