Boston, Massachusetts-based self-driving car startup nuTonomy made headlines in 2016 when it launched the world’s first self-driving taxi service.

Albeit a trial only, the service, which covered a few streets in Singapore, provided an early look at a likely future where self-driving cars run around town offering rides. nuTonomy says a self-driving taxi service is much closer than most people think. In fact, the company says it hopes to launch the first fully functional service in Singapore as early as 2018.

nuTonomy’s progress has caught the eye of automotive supplier Delphi which on Tuesday announced it is acquiring the self-driving car startup in a deal worth $450 million. The move will see more than 100 nuTonomy employees, including 70 engineers and scientists, added to Delphi’s more than 100-member self-driving car development team.

By combining efforts with nuTonomy in Boston and Singapore, where nuTonomy’s operations are currently located, it will mean Delphi will have 60 self-driving cars on the road across three continents by year-end. Other pilot cities include Pittsburgh, Santa Monica and Silicon Valley, and there are more set to join the list.

Delphi showcased a self-driving car prototype based on the Audi Q5 as early as 2015. The prototype successfully traveled from San Francisco to New York City with only minimal driver involvement.

Delphi’s goal is to develop a self-driving system that can be supplied to the major automakers or other customers such as firms looking to offer a self-driving taxi service. One of the customers might end up being ride-sharing company Lyft which is already working with nuTonomy to learn about and optimize the user experience of a self-driving taxi service.

As competition in the self-driving space intensifies, we’re seeing more and more consolidation among the major players. For example, Intel recently acquired sensor technology firm Mobileye; Volvo joined up with supplier firm Autoliv; and General Motors has bought a $500 million stake in Lyft.

Clearly the first group to develop a safe and reliable self-driving system will have a huge advantage as the actual service is hard to differentiate, meaning latecomers’ efforts may be futile.