How Similar are the Safety Systems of Mobileye and Nvidia?

The specialist for autonomous driving, Mobileye, sees an imitation in an Nvidia product. A published document underlines the similarities.

The Israeli company Mobileye has earned much recognition in the field of autonomous driving. Not only do they produce chips and sensors for autonomous driving, Mobileye also produces important software. This has attracted many business partners and in March 2017, they were taken over by Intel.

Nvidia is one of the industry’s leading suppliers of chips and software for autonomous driving. In cooperation with the supplier ZF, Nvidia has created the fastest computer currently available in the automotive sector and is also a leader in the field of artificial intelligence.

Some time ago Mobileye launched a system on the market that is intended to change autonomous driving in the long term. This is called Responsibility Sensitive Safety, or RSS. This is an artificial intelligence that acts as a safety protocol. The motto of the AI is to review decisions according to common sense. It thus accompanies the control of the autonomous vehicle and intervenes in an emergency.

Nvidia’s similar concept is Safety Force Field and is abbreviated to SFF. The head of Mobileye, Amnon Shashua, raised the allegation of plagiarism against this system in his paper “Innovation Requires Originality. He even spoke of a copy “in green and black”, which are the colors of Nvidia. According to Shashua, the system is only a development of its own. When Mobileye presented the system to the public, the industry wanted to be involved in the development. One of the cooperation partners at that time was Nvidia, who later quit. One year later Nvidia released its SFF system.

Shashua sees examples for the similarity of the systems in the evaluation of dangers, like the distance: with Nvidia one would speak of dimensions. A danger threshold that is precisely defined in RSS is called a critical moment in SFF.

Since the development of the technology costs a lot of money, some companies use “shortcuts” to the goal. There are just a few such processes. For example, Apple sued two former employees who were supposed to have copied data in order to overtake the competition. Another well-known case is that of Anthony Levandowski.  Levandowski took data about the Google’s lidar, which has now been thrown onto the market, to Uber. Uber had to hand over 245 million US dollars in shares to Waymo.