It sounds like something out of science fiction, but artificial intelligence (AI) technology is increasingly being used in connected cars.
According to a new report from IHS, in-vehicle infotainment systems and advanced driver assistance systems (ADAS) are growing in complexity and penetration, creating a need for hardware and software solutions that support AI.
“An artificial-intelligence system continuously learns from experience and by its ability to discern and recognise its surroundings,” explained Luca De Ambroggi, principal analyst for automotive semiconductors at IHS Technology. “It learns, as human beings do, from real sounds, images, and other sensory inputs. The system recognises the car’s environment and evaluates the contextual implications for the moving car.”
IHS expects AI-based systems in automotive applications to become standard in new vehicles within the next five years.
Unit shipments of AI systems used in infotainment and ADAS systems are projected to rise from 7 million in 2015 to 122 million by 2025.
IHS’s analysis also showed that the attach rate of AI-based systems in new vehicles stood at 8% in 2015, most of which were focused on speech recognition. That number is forecast to rise to 109% in 2025, with multiple AI systems of various types installed in many cars.
In particular, AI systems will be used for:The human-machine interface for infotainment systems, incorporating speech recognition, gesture recognition (including hand-writing recognition), eye tracking and driver monitoring, virtual assistance and natural language interfaces.ADAS and autonomous vehicles, including camera-based machine vision systems, radar-based detection units, driver condition evaluation, and sensor fusion engine control units (ECUs).
Some elements of AI are already available in vehicles today. For example, while most speech recognition technologies rely on algorithms based on neural networks running in the cloud, the 2015 BMW 7 Series has embedded hardware that can perform voice recognition in the absence of wireless connectivity. And in ADAS applications, Tesla says it has implemented neural network functionality in its autonomous driving control unit.