If carmakers really want to offer self-driving vehicles to the market, they should not just focus on mechanical engineering but also on software design, Maarten Sierhuis, a computer scientist director of Nissan’s Silicon Valley research center, told Bloomberg in an interview.
Sierhuis remarked that carmakers should build cars that are “deliberative” in assessing road conditions, instead of reactive ones. He noted that this feat requires artificial intelligence.
Nissan is targeting to market self-driving vehicles by 2020 or sooner, and to achieve that goal, it is now developing software that could read and filter sensor data like a human brain does, Sierhuis said.
He told Bloomberg that auto industry needs to a shift from thinking that a car is a physical, mechanical system into replicating autonomous systems with software.
Nissan as well as other car and parts makers like Toyota Motor Corp. and Daimler AG, and tech giants like Google Inc. are making a great push into researching systems that could make driving partly or fully automatic.
Self-driving vehicles are believed to help reduce accidents and congestion and trim fuel use. Self-driving vehicles would also allow people to use the time in-transit for activities other than driving.
Although safety regulators in the United States don’t deny the potential benefits for improved safety from self-driving vehicles, there is a need to address certain regulatory and legal issues like who would be liable in case of accidents.
US President Barack Obama has underscored government-supported research on technology being developed by carmakers that would enable vehicles to communicate with each other to cut traffic jams and accidents.
Sierhuis, known for designing software for NASA space missions, remarked that automated vehicle features like adaptive cruise control, blind- spot monitoring and lane-keeping assistance aren’t enough. [source: Bloomberg]