Turing wants to produce 10,000 fully autonomous electric cars
A self-driving electric car uses artificial intelligence imaging and scanning to drive on highways and in cities. Japanese electric car developer Turing has unveiled its driverless electric vehicle, the vehicle concept of which allows AI to make driving decisions while cruising the road.
The onboard camera collects the views it receives while driving, in addition to sensors and dynamic maps of the vehicle’s design. Based on the collected information, the car independently operates the steering wheel and drives independently on public roads without the help of passengers.
Images courtesy of Turing
The move towards AI has always been the concept of the car, and in 2022 the technology team was able to amass a driving database of 500 hours for its AI-driven system. For the car design, Turing is collaborating with design group Nichinan to adopt Stable Diffusion, deep learning and text-to-image modeling.
The team decides to combine static imagery with digital modeling, CG rendering, full color 3D printed models, driving animations and augmented reality data. The design result is a bean-shaped model that tilts forward for a more rounded look, swings up the doors and splits the windshield into three views.
“By combining artificial intelligence and vehicle design, this project, which brought together experts in human intelligence and digital labor, demonstrates the amazing power and speed of AI design,” said Yoshiichi Saruwatari, general manager of the Design and Engineering Division and director of Nichinan. Turing plans to complete a small production facility in 2023 with the order of 100 units and 10,000 hours of driving data collected, and to reach a production of 10,000 fully autonomous electric vehicles by 2030.
Turing plans to beat Tesla with its fully autonomous and self-driving electric car
Creation of a large-scale deep learning model for electric vehicles
While Turing has achieved groundbreaking achievements, the design team recognizes the amount of time they must invest to realize their project’s full potential. You will be informed that fully automated driving requires human recognition and judgement.
They believe that in order to realize fully automated driving, it is necessary to create a large-scale deep learning model for their electric car. “We have already built a significant amount of driving databases, but we plan to expand them exponentially in the future. In the future, we will create a system that can collect driving data from vehicles sold by Turing,” says the design team.
The team combines static images with digital modelling, CG rendering, 3D printed scale models and AR data
Turing is aiming for mass production of its fully autonomous electric car and even claims to achieve its mission to overtake Tesla. “If hardware and software get along well, we can build better cars from Japan. I will put this wonderful misunderstanding into action with my friends,” says Turing co-founder and CEO Kazunari Yamamoto. “I decided that my next challenge in life would be to start an automaker that would outperform Tesla.”
Shunsuke Aoki, co-founder and CTO of Turing, echoes Yamamoto’s statement of bringing his fully autonomous electric car to life. “I have devoted much of my life to self-driving cars and academic research. When I returned to Japan, I thought about the question, “What kind of work can I be proud of for future generations?” And I decided to set up mass production of fully automated electric vehicles with Mr. Yamamoto,” he says.
