Long before robotaxis became a headline, a small fleet of driverless shuttles was already carrying office workers between a Rotterdam metro station and a business park. That system, built by Utrecht-based 2getthere, opened to the public in 1999 and is still running. In October 2026 its third-generation vehicles welcomed their one millionth passenger, and the company that built them now has a new owner, a new headquarters and its largest project yet: a 390-vehicle pod network planned for Mumbai.
Two flavours of driverless transit: GRT and PRT
2getthere builds automated people movers that run on dedicated lanes or guideways rather than in open traffic. It offers two system types. Group Rapid Transit (GRT) uses larger shuttles shared by many riders, while Personal Rapid Transit (PRT) uses small pods that carry a single party directly to its destination on demand.
The flagship GRT system is the ParkShuttle at Rivium in Capelle aan den IJssel. According to 2getthere, six fully electric autonomous vehicles, each holding 22 passengers, now carry more than 1,850 people a day on a 1.8 kilometre route with six stops, linking the business district to Rotterdam’s Kralingse Zoom metro station. The company’s Rivium project page says the line has carried more than 5.4 million passengers since 1999 and that the vehicles handle crossings where other road users cross the track.
The best-known PRT installation is at Masdar City in Abu Dhabi, opened in 2010. The Masdar project page describes 10 electric six-person pods on a 1.4 kilometre guideway, with more than 2.7 million passengers carried and 99.6% system availability over 15 years of daily service.
The software that runs the fleet
The vehicles are only half the story. In an August 2026 interview published by 2getthere, team lead Karthik Venkatapathy describes a two-layer architecture. On board, vehicle control software called FORCES, together with perception systems and a safety controller, handles immediate safety decisions such as reacting to obstacles. Above it, a fleet platform called TOMS (Transit Operations and Management System) allocates vehicles to demand, plans collision-free and deadlock-free routes, reroutes when conditions change and keeps passenger information up to date.
That separation matters as systems grow. Venkatapathy says the Mumbai project required TOMS to scale from supporting around fifty vehicles to fleets of up to four hundred, which forced a rethink of how alerts reach human operators so they are not overwhelmed. He also notes growing interest from organisations that run their own autonomous vehicles and want a proven fleet management layer, and says AI tools are helping the team develop and test new features faster.
From ZF subsidiary to Indian ownership
2getthere has changed hands twice in seven years. In March 2019 ZF Friedrichshafen announced it had bought a 60 percent stake, citing more than 100 million kilometres of autonomous mileage across the company’s passenger and cargo systems and reliability above 99.7 percent. At that point the Rivium and Masdar systems together had carried more than 14 million people, ZF said, and 2getthere employed about 60 staff in Utrecht.
The automotive supplier’s ambitions cooled. In December 2023 ZF said it would stop pursuing complete autonomous transport systems, including shuttles and fleet management, because the market was developing more slowly than expected and up-front investment was no longer justified. Existing GRT projects such as Rivium would continue, and US partner Oceaneering kept its licence to produce and sell the GRT shuttle.
In January 2026 advisory firm Hampleton Partners announced that India’s Sai Green Mobility had acquired 100% of 2getthere’s shares from ZF after an international bidding process. The Indian Express reported that the acquisition took place in the summer of 2025, alongside the Mumbai project changes described below.
Mumbai: the 390-pod test of scale
The new owner’s interest is tied to a pod taxi network in Mumbai’s Bandra Kurla Complex (BKC), a dense business district. According to 2getthere’s BKC project page, the plan calls for 390 autonomous, electric six-person PRT vehicles on 8.8 kilometres of elevated guideway, operating on demand either individually or in platooning mode, with operations planned for 2027. The company’s About page projects more than 100,000 passengers per day once service starts.
The route to that point has not been simple. The Indian Express reported that UK-based Ultra PRT, the technology partner on Sai Green Mobility’s winning bid, parted ways with the concessionaire in July 2024. The Mumbai Metropolitan Region Development Authority (MMRDA) approved 2getthere as the replacement in September 2025 after a technical committee review, using a tender clause that lets the technology partner change during implementation. As of that June 2026 report, barricades had gone up near the BKC Metro station and soil testing had begun.
What it means
2getthere’s record shows that driverless transit can work as everyday public transport when the operating environment is controlled. Dedicated lanes, known stops and predictable demand let the company sidestep many of the hardest problems that open-road robotaxis face. Its Rivium milestone is a reminder that “autonomous” does not have to mean experimental.
For large campuses, the model is directly relevant. In a November 2026 interview on the US market, business development representative Cameron Gieda names airports, large healthcare campuses, universities and stadiums as the strongest fits, and argues that such systems can compete with light rail at a fraction of the cost. He also says the main US barriers are usually funding, real estate and the number of agencies involved, not the technology itself. For hospital and university leaders, that points to a practical question: whether a fixed-route, on-demand shuttle could link parking, transit stops and buildings more reliably than buses.
The open question is scale. Moving from a six-vehicle line and a ten-pod network to a 390-vehicle system in one of the world’s busiest cities is a very different challenge, and it will test both the TOMS fleet layer and the new ownership structure.
Conclusion
2getthere has spent more than 25 years proving that driverless shuttles can run quietly and reliably on dedicated routes. Its next chapter depends on Mumbai. If BKC opens as planned in 2027, the company will have shown that its approach can grow from a business park shuttle into mass transit for a major city.
Sources
- 2getthere: The world’s first self-driving transit system
- 2getthere: One million passengers. One proven autonomous transport system
- 2getthere: Rivium project
- 2getthere: Masdar City project
- 2getthere: Bandra Kurla Complex project
- 2getthere: About us
- 2getthere: The invisible intelligence behind autonomous mobility
- 2getthere: Autonomous transit in the US: turning experience into opportunity
- ZF: ZF acquires mobility provider 2getthere (2019)
- ZF: ZF strategically adjusts its shuttle business (2023)
- Hampleton Partners: Hampleton Partners advises ZF Group on the sale of 2getthere
- The Indian Express: The Heathrow pod taxi company once backed Mumbai’s project. Then it walked away