A Hungarian artificial intelligence company is angling to make a name for itself in self-driving cars. AImotive, changed its name fromAdasWorks and opened a United States office as it expands.
Started in 2015, AImotive has grown from 15 employees to 120 people working on Level 5 architecture, including a hardware-agnostic software suite, an AI training toolkit, and neural network hardware IP.
The development of Level 5 self-driving vehicles still faces both technical and legal challenges. In order to be classified as L5, a vehicle needs to be able to drive itself completely autonomously, under any circumstance a daily driver might encounter, including inclement weather and unexpected obstacles. The levels were established by the National Highway Traffic Safety Administration, which later adopted the Society of Automotive Engineers’ slightly different but comparable five-level system. So far, Tesla and Google seem to be on the forefront of the effort to make vehicles with level-five autonomy.
AImotive’s expansion includes the office in Mountain View, Calif., the same city where Google’s autonomous car test bed is located.
AImotive’s technology stack includes the ability to gather, analyze, and learn from sensor data. Its segmentation tool recognizes up to 100 different objects a car may encounter. AImotive will market several globally scalable aspects of this stack, including the Location Engine (a mapping system which uses 3D landmark point data as well as GPS positioning), the Motion Engine (which tracks and predicts moving objects), and the Control Engine, which integrates with and manages the car’s basic and auxiliary functions.
The above are all included in the aiDrive stack. AI motive also offers aiKit, a training tool, and aiWare, their custom hardware design for embedded solutions. It incorporates high bandwidth and low latency Neural Network (NN) computation in a low-power package. It can be integrated with GPU, FPGA, or embedded systems, contributing to AImotive’s goal to be processor-agnostic.
AImotive has raised $10.5 million in funding since its launch in 2015, as well as working with other artificial intelligence projects like CEVA Inc.’s autonomous vehicle vision system.
