SAN FRANCISCO (Reuters) - Hungarian autonomous vehicle technology start-up AImotive, which has changed its name from AdasWorks, said on Monday it had expanded to Silicon Valley, opening the door to becoming an acquisition target.
The company, which has raised $10.5 million in seed and Series A funding from investors including Tier One supplier Bosch and graphics chip maker Nvidia Corp (NVDA.O), opened an office in Silicon Valley.
AImotive is developing software that combines artificial intelligence with a low-cost system for gathering information about the conditions around a vehicle. A production version of the system for enabling autonomous driving will be ready in two years, executives said.
“If you just see the statistics, we will probably be an acquisition,” Chief Executive Officer Laszlo Kishonti told Reuters.
In the highest-profile deal to date, General Motors Co (GM.N) bought self-driving start-up Cruise Automation for $700 million in May, sending a ripple effect through Silicon Valley’s fledgling autonomous vehicle technology companies.
“I think only a minority of (automakers) will be able to execute this plan all alone,” Kishonti said. “That’s why specialists like us who are hardware-agnostic and neutral to different clients can help them with software technology.”
Kishonti said AImotive, which plans to open offices in Japan and China next year, used off-the-shelf components to keep costs down. The company has grown from 15 engineers to 120 engineers and researchers.
AImotive’s system relies on six to 12 cameras in each vehicle that cost about $15 apiece. In contrast, the original laser-based Lidar sensor made by Velodyne that twirls atop Alphabet Inc’s (GOOGL.O) Google self-driving cars costs about $75,000, although Lidar manufacturers are working to slash the solid-state device’s cost to about $100 by giving it fewer moving parts.
Artificial intelligence, driven by faster and more powerful chips, can also substitute for the high-definition, three-dimensional maps that many in the industry say are a key ingredient to allow a car to drive itself.
“If you wanted have the full United States in 3D, that would be thousands of terabytes,” Kishonti said. “I think it’s not really feasible to have a global product where the whole world is stored in your trunk and needs to be updated.”
Reporting by Alexandria Sage; Editing by Lisa Von Ahn