Mobileye, a company owned by Intel, is actively expanding its autonomous vehicle initiative and intends to begin testing vehicle fleets in at least four additional cities in the coming months: Detroit, Paris, Shanghai, and Tokyo.

During a presentation at the virtual 2021 CES technology event on Monday, Mobileye’s president and CEO, Amnon Shashua, stated that the company also hopes to initiate testing on public roads in New York City, pending the necessary regulatory approvals.

This announcement of expansion, alongside information regarding a new lidar System on Chip currently in development and slated for release in 2025, demonstrates Mobileye’s commitment to the commercialization of self-driving vehicle technology and making it widely available.

According to Jack Weast, a senior principal engineer at Intel and Vice President of Automated Vehicle Standards at Mobileye, the choice of these cities and countries is determined by two key considerations: the presence of customers and the prevailing regulatory landscape.

“This is why we are deploying vehicles in Detroit, rather than Silicon Valley, as the major automotive manufacturers are located in Detroit,” Weast explained in an interview Monday. He further noted that Peugeot and Renault are based in Paris, while Toyota and Nissan operate in Japan. “Selecting these cities allows us to position the vehicles close to our customers, providing them with direct experience of the technology, as we anticipate our OEM customers will remain vital to our business, even as we offer a complete self-driving system.”

The company reports that a test fleet is already operating in Detroit. Mobileye initially launched a test fleet in Jerusalem in 2018, followed by another in Munich in 2020.

Mobileye is pursuing a comprehensive three-part strategy for the development and implementation of automated vehicle technology. This strategy integrates a complete self-driving system – incorporating redundant sensing subsystems utilizing cameras, radar, and lidar – with its REM mapping system and a rules-based Responsibility-Sensitive Safety (RSS) driving policy. Mobileye’s REM mapping system leverages data gathered from approximately 1 million vehicles equipped with its technology to create detailed high-definition maps, which can then be used to enhance both ADAS and autonomous driving systems. Shashua indicated that Mobileye’s technology can now automatically map the world, tracking nearly 8 million kilometers daily and accumulating almost 1 billion kilometers of data to date.

Shashua stated on Monday that this approach will enable the company to efficiently launch and operate commercial robotaxi services, as well as integrate the technology into consumer vehicles by 2025.

Mobileye has historically been a leading provider of computer vision sensor systems designed to prevent collisions within the automotive industry. In 2018, the company broadened its scope beyond simply supplying components to also operating robotaxis, and is now focused on bringing autonomous vehicle technology to passenger cars by combining its computer vision expertise with the new lidar SoC it is developing in collaboration with Intel.

Mobileye currently collaborates with Luminar for lidar supply in its robotaxis. However, the company shared further details about the lidar SoC, which it anticipates will be ready for passenger vehicles by 2025. Neither Shashua nor Weast commented on whether the partnership with Luminar would be discontinued once its own lidar SoC is available.

The lidar, which will be manufactured using Intel’s specialized silicon photonics facilities, is significant because Mobileye is well-known for its camera-based technology. Despite this, the company is not abandoning its camera-first approach. Shashua clarified that Mobileye believes the most effective technological and business strategy is to develop a camera-based system and then incorporate lidar and radar as supplementary layers for redundancy.

“The core concept is to have this camera subsystem,” Shashua said. “Because it’s camera-based, it’s available at a consumer price point, enabling scalable thinking. This scalability is key to long-term sustainability until level four automation becomes commonplace.”

Shashua cited the company’s long-term agreement with Geely Auto for advanced driver-assistance systems as an illustration of how a camera-first approach can be adapted over time. Lidar and radar can be added to provide enhanced automation capabilities as the market matures.