Merli told us that interest in autonomous cars is growing rapidly in Asia as well. "Vision based solutions, sensors, imaging processors, radars, LIDARs: car makers and system makers are increasingly eager to adopt and integrate such products in their solutions, either stand-alone or integrated with other Control Units," he said. "Autonomous driving is supposed to be the fastest growing market in automotive (and Asia and China will be leading it) in the next years."

Eduardo Merli Merli: EyeQ5 will come at the right time--with superior performance, towards achieving autonomous driving.

With engineering samples of the EyeQ5 expected only by first half of 2018, is STMicroelectronics too late? Probably not. According to IHS, STMicroelectronics has a broad- based portfolio and a presence in every growing automotive domain of the market. That helped the company maintain its revenue despite being affected by the lower exchange rate of the Euro against the U.S. dollar--it dropped by 20%--in 2015. STMicroelectronics’ automotive revenue declined 2% year-on-year, mostly due to the change in exchange rates.

"EyeQ technology is extremely successful in the market, EyeQ3 is entering in its peak production and EyeQ4 is being sampled to customers. Definitely the solution, together with other ST technologies (radars, V2X, etc.) responding to the market and customer requests for next year designs, EyeQ5 will come at the right time--with superior performance, towards achieving autonomous driving," said Merli, defending the company's product roadmap.

The EyeQ5 adds other sensor data over the vision-focused approach in the previous versions. According to Merli, the product leverages Mobileye's vision processing expertise--it hosts dedicated processors and cores to implement sophisticated algorithms. "The higher processing capability offered by the technology and the architecture of the device allows [you] to process a large number of information sources and then implement the appropriate actions. So EyeQ5 from one end continues and improves the path of Mobileye superior vision based processing capability and from the other end offers the possibility to integrate and process other information sources," added Merli, explaining the approach toward sensor fusion that is being taken by other companies as well.

Technical challenges remain. Merli said that the prerequisite to full autonomy in vehicles is a powerful computing platform and engineers will focus artificial intelligence, adaptive behaviour, multi-sensor and connectivity integration and security. "Standardization will also play a key role (on connectivity for instance)," he said.

Merli sees challenges beyond the engineer's desk, however. "The key issue will probably be legislation (i.e. ethic, liability, business models, new car ownership concept) along with other issues, which have to be resolved to allow massive deployment," he explained.

<< Previous page: Increasing basic autonomy