AI and automobiles

AI is poised to assume a greater role in ensuring driving safety as car manufacturers look to incorporate the technology into the production of automobiles. Subaru and Dell are partnering to combine AI and high performance storage that will aid in the development of the Japanese car maker’s well-received Eyesight driver assist technologies. Similarly, Microsoft and automotive supplier Bosch say they will partner to develop generative AI to improve automated driving.

Subaru will benefit from access to Dell’s Powerscale data storage tech that allows the car maker to store 1,000 times more files than before to aid in next-generation Eyesight development solutions. Subaru can now improve AI image analysis by easily accessing stored files on Dell PowerScale systems deployed in data centers across Subaru Lab and Tokyo offices, something that previously wasn’t available. Subaru’s Eyesight system uses stereo cameras to provide driver assist features such as pre-collision breaking and adaptive cruise control.

“Subaru is driving massive innovation through data to give motorists an extra set of eyes and an extra foot on the break,” said Arthur Lewis, president of Dell’s Infrastructure Solutions Group.

“While requirements for systems and storage are ever-changing, we are confident the Dell PowerScale is up to the task as the underlying infrastructure for Eyesight driver assist technology, allowing us to continue our AI initiatives to improve driver safety,” says Takashi Kanai, deputy chief of Subaru Lab and manager of the car maker’s ADAS Development Department, PGM Engineering Division.

Dell stock has soared to new highs on the back of demand for information storage technology equipment to handle AI work. Dell’s infrastructure unit reported revenue of $9.33 billion in the fiscal fourth quarter, exceeding estimates.

Microsoft and Bosch, meanwhile, hope generative AI can lead to a new level of driver safety. “The expectation is that generative AI will enable vehicles to assess situations and react accordingly and in this way, keep road users even safer,” said Bosch chairman Dr. Stefan Hartung at the company’s just concluded AIoT conference in Berlin.

While current driver assist technologies can detect animals, people, objects and vehicles, it’s hoped that future generative AI could determine situations that may cause an accident. Generative AI trains on massive amounts of data and may be able to draw conclusions from this information. These conclusions could be communicated directly to the driver as a warning or instructions to initiate appropriate driving maneuvers such as braking.

The ability of car companies to gather information that might support AI learning about drivers – using data relayed to car makers via OTA links – is being called into question as the U.S. government probes whether Chinese vehicle imports pose a national security risk. The White House said the Commerce Department probe is necessary because “connected cars “collect large amounts of sensitive data on their drivers and passengers, and regularly use their cameras and sensors to record detailed information on U.S. infrastructure.”

The U.S. probe will include autonomous vehicles over concern that these vehicles could be piloted or disabled remotely. Such data-gathering capabilities are common to many new vehicles regardless of country of origin, so the impact the Chinese auto probe remains to be seen.

Compounding the difficulty are plans by Chinese auto companies like BYD to build factories in Mexico which would largely circumvent U.S. import restrictions. The security of data collected by car companies across the board is a growing area of concern. Most car companies receive a dismal rating by analysts when it comes to data security.