Autoliv Research to reveal ‘Learning Intelligent Vehicle’ at CES 2017

Autoliv, the global leader in automotive safety systems, is to launch a learning, intelligent vehicle (LIV) system. The system is developed by Autoliv Research to facilitate collaboration and shared control between driver and vehicle. This in order to attain trust and facilitate required safety in autonomous driving. The LIV-system is to be demonstrated at CES in Las Vegas on January 4.

The first version of LIV, the Learning Intelligent Vehicle, utilises Artificial Intelligence (AI) technology, and coming versions will incorporate on-line and off-line machine learning, utilizing a wide array of data sources and developing an individual relationship to the driver. LIV, pronounced {'li:v} or ‘leave’, literally means ‘life’ in Swedish and is a common Swedish first name as well as part of Autoliv's own life-assuring brand name.

“With this research platform, Autoliv Research aims to solve what truly proactive safety is all about. With LIV, we have found a way to combine our history of saving lives and our leading technology within active safety to accelerate the safe development of automated vehicles,” said Ola Boström, vice-president (Research) at Autoliv.

LIV 1.0

The first version of LIV is equipped with Autoliv technology readily available today, connecting all the different technologies via an AI-based ECU, which communicates and acts on all available data. Data sources include what is derived from the driver, the car, its surroundings, as well as from other vehicles and external data sources.

In both autonomous driving situations as well as with the driver in control, LIV complements the driver, monitoring and analyzing all aspects surrounding the vehicle and safeguarding the situations in close cooperation with the driver.

LIV also provides assistance based on the traffic around the car as well as the preferences of the driver, keeping the driver in the loop by allowing for two-way communication as well as interpreting the driver's behavior. In forthcoming versions, LIV will also incorporate machine learning technology, providing the necessities to learn the unique characteristics in the relation between the individual driver and vehicle.

The main reason for promoting the LIV research platform, incorporating and testing new and existing technologies on a continuous basis, is to develop the necessary conditions to be able to unleash autonomous driving and AI technology on a broad scale – while maintaining safety and providing a better driving experience. The accelerated innovation and technology development within autonomous driving and automated transport solutions in turn requires for more of a so called human centric approach – where shared control and two-way trust between human and smart vehicle is crucial for a safe autonomous development.

Ola Boström, who is in charge of the LIV project, explains further: “The more advanced machine learning and AI technology becomes, the more important it is to have a human centric approach – understanding how we humans and smart machines collaborate as a joint cognitive system. As vehicles become safer, in a sense we become worse drivers. We no longer need to control the vehicle unless it needs or wants us to. With LIV, we have turned this table around, with a system that trusts you to drive and works together with you, and not the other way around.”

Autoliv’s CES 2017 showcase

Mono Vison, Stereo Vision, Night Vision: Autoliv's custom algorithms support adoptable and scalable vision system solutions in the day or at night. With Mono Vision, there is more safety for the driver and vulnerable road users. Even if the driver is distracted, the car will be there to avoid the crash. Stereo Vision can detect classified objects (pedestrians, cars, trucks), lane markings, traffic signs and road edges.

Night Vision sees objects in darkness – through smoke, fog and in challenging lighting conditions. Based on Far Infrared technology, the sensors are not impacted by light and can clearly see beyond blinding headlight and sunlight glare. Additionally, Animal Detection and Dynamic Spotlight are part of this revolutionary safety system.

Radar and Sensors: As one of the leading automotive radar system suppliers, with about seven million sensors delivered in 2016, Autoliv is providing high-resolution radar systems for the ever-increasing standards of the global 5-star NCAP car safety programs and supporting highly automated driving capability. The 77 GHz radar family is being extended from the corner radars currently in production to forward looking and highly innovative satellite radars, enabling radar based object detections all around the car.

Automated Emergency Braking (AEB) system: Utilises fused radar and camera input to ensure high confidence in braking; camera-only information can be used to brake for vulnerable road users.

zForce DRIVE Intuitive Steering Wheel and RoadScape GPS Electronic Horizon: HMI-based innovation designed to keep the driver in full control, demonstrating the seamless transition between manual and automated driving. By using touch technology and air gesture control, zForce can anticipate and react to human actions to adjust to all road surfaces and weather conditions.

Accurate and reliable positioning and localisation are critical needs for today's vehicles with advanced driver assistance systems (ADAS) and active safety, and even more essential for tomorrow's transportation revolution in autonomous driving systems. Autoliv has announced its sixth generation of GPS – Horizon software provides context of the environment of the road such as road signs, curves, hills, and tunnels greatly improving the performance of line-of-sight sensors used in advanced driver assistance systems.