ranjit | Feb 19, 2018 | 0
Autonomous drive test in fast lane
By TA News Bureau:
Global semiconductor giant Intel’s $15.3billion acquisition of Israel’s Mobileye is a defining development. It will accelerate electronic driverless systems that would propel autonomous driving to new heights. The deal will catapult Intel into direct competition with rivals Nvidia and Qualcomm Assured of the front rank position in the specialist emerging automotive suppliers’ league, the acquisition is set to stir up the market for advanced driver assistance systems and autonomous vehicles that is set to grow from about $3 billion in 2015 to $96 billion in 2025 and $290 billion in 2035. Analysing the situation Akhilesh Kona, Senior Analyst with Automotive Semiconductor of IHS Markit Technology, says in an interview with Tyre Asia that Inteisition would lead to a quantum jump in the efforts at enlarging the scope of automated driving
Prior to the acquisition of Israeli startup Mobileye, Intel — the world’s largest computer chip-maker — had announced partnership with BMW to develop autonomous cars around 2020. Mobileye on the other hand had partnerships with many Tier-1s and car OEMs for automated driving with its processors and machine vision algorithms.
“With this acquisition Intel will have a broader access to the autonomous car supply chain with its existing connectivity solution and Mobileye’s solution for machine vision, Road Experience Management (REM) and processors for automated driving,” says Akhilesh Kona, Senior Analyst, Automotive Electronics & Semiconductors at IHS Markit Technology. He is responsible for market research on sensors and semiconductor content, including various technology trends in the field of Advanced Driver Assistance Systems.
In an interview he told Tyre Asia that in addition to Mobileye, Intel’s other past acquisitions including Altera (automated driving hardware platform), Yogitech (functional safety expertise), Movidus and Nervana (artificial intelligence) and Arynga (over-the-air updates) will position the chip-maker as a one-stop-shop for infotainment and automated driving solutions.
Elaborating on his observation that the merged entity will be able to address the core elements of self-driving cars such as sensor fusion and front-view camera that would ensure greater safety, Kona said that object recognition, sensor fusion, path planning, localisation, connectivity and telematics are some of the key building blocks to enable a safe and reliable driving policy in autonomous cars.
“Mobileye is the current market leader in front-view camera modules, including object recognition, and it has strong product roadmaps for sensor fusion software and hardware platforms called EyeQ5.”
In 2016, Mobileye announced its design wins from five OEMs for its L3 type of automated driving and is currently developing its localisation (REM) with 3 major OEMs. The merged entity will focus on the key building blocks required for automated driving system through Mobileye’s expertise in machine vision and sensor fusion, including software, artificial intelligence, hardware platforms (EyeQx processors) and REM
On Intel’s expertise in sensor fusion through artificial intelligence (Movidus & Nervana), functional safety (through Yogitech) and hardware platforms (Intel and Altera), the acquisition would help Intel to soar into market leadership.
Referring to the technological developments that are currently on the anvil in automated driving, Kona said Light Detection and Ranging (LIDARs), sensor fusion (domain controllers), localisation, artificial intelligence and connectivity and telematics are some of the technologies that are currently under development to implement automated driving.
“These technologies are mostly developed by startups or spin-offs of research and development groups from research institutes and universities,” explains Kona, who has a Master’s degree in Electrical Engineering from Deggendorf University of Applied Sciences in Germany.
He affirms that Intel’s acquisition of Mobileye is a major change for automotive supply chain since Qualcomm’s acquisition of NXP. “Sensor fusion, front-view cameras and telematics and connectivity are some of the fastest growing segments in automotive sector. The combined system revenues for sensor fusion and front-view camera modules will reach $10 billion by 2022.”
He said Qualcomm-NXP merger targets mostly hardware platforms (processors, sensors, wireless ICs, etc.), but the Intel-Mobileye entity in addition, would also provide software on top of hardware platforms (processors, wireless communication ICs, etc.).
“This will put Intel in a prevailing position and poses a tough competition to Qualcomm-NXP merged entity,” he observed.
Commenting on reports that these acquisitions are the way forward in boosting technology development in automated driving systems, Kona is sure that these will accelerate technological developments. “Going forward, acquisitions and partnerships will accelerate the development and deployment of automated driving solutions.”
Autonomous functions are majorly supported by machine vision, artificial intelligence. Artificial intelligence features are new to traditional automotive supply chain, but the scene is changing, he notes.
“Most of the traditional automotive suppliers are gaining access through such technologies through acquisitions such as GM-Cruise, Delphi-Ottamatika, etc and through partnerships such as NVIDIA-ZF, Bosch-ZF, etc.”
For example, at the semiconductor level, NVIDIA traditionally has come from the gaming industry which gives them an edge over traditional automotive chip suppliers like Renesas, NXP, Infineon and Texas Instruments in developing graphical processor units – which are important to implement algorithms for sensor fusion, machine vision and artificial intelligence.
On the other hand, hardware platforms from traditional automotive chip suppliers implement functional safety tasks (ISO 26262) associated with automated driving, Kona explains.
With Intel’s acquisition of Mobileye, the autonomous vehicle market would see a huge acceleration in driverless cars.