Building trust in autonomous vehicle safety
Societies around the world are currently facing a plague, the effect of which can be seen on our roads every day. Traffic-related deaths have climbed to 1.4 million annually, and that number could double if we don’t make real changes in the way we design cars. For the past 22 years, I have worked at Autoliv to battle this deadly trend by putting technological research into action.
Autoliv is the world’s largest automotive safety supplier, working with car manufacturers to make driving safer around the world. Together with our ecosystem of partners, which includes Ericsson and MIT researchers, we are driving forward the technology that will not only make autonomous vehicles safe for our roads, but also developing approaches which build trust in self-driving technology.
Our ecosystem panel at Consumer Electronics Show 2018
I recently took part in a panel discussion in Las Vegas as part of CES 2018. Together with Bryan Reimer, Ph.D., Research Scientist in the MIT AgeLab and the Associate Director of The New England University Transportation Center, and Erik Ekudden, CTO Ericsson, we were able to present ideas about autonomous vehicle safety from the point of view of a technology supplier, a researcher and a connectivity expert.
It is exactly this type of collaborative ecosystem approach that will lead the way to safer roads and ultimately the emergence of the self-driving vehicles. Too often, industries develop technology in proprietary silos. This approach may be good for speed, but it is detrimental to consumers, and ultimately, slows down the adoption rate for new technologies. With a collaborative approach, we hope to create widely accepted standards that get closer to what people really want and need in their connected, automated cars.
Perfecting each link strengthens the chain
During our discussion at CES, the panel examined the changes which need to take place in order to get safe autonomous vehicles on the road. The technologies that makes advanced driver assistance systems (ADAS) possible today are the foundation. Those technologies include radar, lidar, exterior cameras and sensors, strong computing power and software that can evolve over time.
But today, there is still fragmentation between the components that will make up the operational system of the truly autonomous, connected cars of the future. The technology behind autonomous drive, driver-sensing applications and vehicle connectivity still need to be perfected on an individual basis.
This process is an evolution, not a revolution – it will not happen overnight. As each of these components are complex in themselves, the way forward is to start with a vision of a unified system, and perfect those individual components. Over the next five to ten years, the fusion of all these aspects should come together in a complete package.
Moving towards human-system integration
To increase safety on our roads, simply developing new technology is not enough. A better understanding of how humans and smart vehicles collaborate will make the adoption process smoother, and more successful in the long run. Part of what I want to achieve is taking one step further from HMI (human-machine integration) to HSI (human-system integration).
Before we advance to full automation, many drivers will be more comfortable with a vehicle that acts as a co-pilot, as opposed to the captain. In the past, the focus has been on teaching robots how to drive. Now we are able to begin using AI to create a deeper experience, building on the established path-planning technology.
Integrating interior cab sensing with the external sensors can augment human decision making based on the capabilities and status of the operator at any given time. If the driver is fully focused, the automated car can step back. But when the driver takes a call or grows tired, the operational system can take a decisive role.
Beyond just operation, adaptive automation and driver interfaces based on the entire transport experience will prove to be very valuable. When our cars know what music to play, what temperature the cabin should be, and whether to take the scenic route or the expressway, then drivers will have a truly personalized experience.
Helping the auto industry take a human-centric approach to smart cars
Consumer research shows that people are interested in autonomous cars, yet many are afraid to let go of the wheel. Convincing drivers that automation will make their lives easier and safer will trigger the mass adoption of self-driving vehicles. The collaborative efforts of Ericsson and Autoliv are making the current transition period easier for OEMs.
Together with software company Zenuity, our two companies are developing a cloud-based infrastructure designed to help automakers accelerate the consumer adoption of safe self-driving vehicles. The platform is a “white-label” solution focused on connected car safety using Autonomous Driving (AD), Advanced Driver Assistance Support (ADAS), AI and Machine Learning technologies.
As we make the transition to connected self-driving vehicles, our mobile networks, AI, IoT and cloud technologies must be as reliable and trusted as the vehicles they support. Ensuring that reliability will require specialized knowledge, experience, and innovation.
With the partnership of Autoliv and Ericsson, we have combined our specialized competencies related to automotive safety, ICT leadership, and an understanding of the pressures on automotive OEMs. I look forward to the day when our combined efforts turn around the trends of road safety with cars that succeed in human-system integration.
This has been a guest post from Ola Boström, Vice President of Research from Autoliv. Find out more information on how Ericsson is working with our partners to make self-driving vehicles a reality here.