Skip to content

How AI will pass its driving test

Nick Gill, Chairman of Global Automotive Sector at Capgemini, considers the benefits of artificial intelligence in cars and across the automotive industry

Artificial intelligence is learning to drive. Within the next 20 years it will take over from humans as the main entity behind the wheel. However, AI in the automotive industry is more than the technology merely passing its driving test. This piece explores not only the impact of AI on driving, but how it will fundamentally alter the way OEMs do business and the laws surrounding road use across the world.

Manufacturers are sitting on a huge amount of information from cars already on the road, but are unable to harness it for sales purposes with insights buried amongst mountains of data. AI will change this

Autonomous vehicles

In thinking about AI in the automotive industry, there is one trend that can’t be ignored: autonomous vehicles. As noted by Tim Cook, the Chief Executive of Apple, developing autonomous vehicles is the mother of all AI projects. Driverless cars, as they are commonly known, will become commonplace on our roads within the next ten to 15 years. Dozens of automotive and technology companies have cars currently in development and, while most public debate has focused on Google and Tesla’s efforts, traditional manufacturers including Audi, BMW, Ford, Hyundai and GM have all entered the autonomous vehicle race. With all these firms trying to get their cars on the road first, at least US$80bn has been ploughed into research in the past three years.

However, this shift to autonomous vehicles will not happen overnight. There will be a lengthy and organic evolution from human-driven cars to those without any human controls. The development of autonomous car technology is measured in five levels, where fundamentally L5 is a car without a steering wheel. Some vehicles, like the new Audi A8, claim to have reached L3 now and we can expect to see L4 cars on the road within the next few years. However, the technology becoming available will only facilitate the move to autonomous vehicles, not immediately make every car on the road driverless. The willingness of humans to step into a driverless car for the first time, coupled with infrequent car purchases will slow the adoption cycle. As with every technology, there will be early adopters, but most consumers will wait until their car needs replacing before assessing whether or not to go driverless.

With AI set to change the automotive industry, there will also be great changes in the way humans use their cars. Instead of having to negotiate the roads and find parking spots, users will be freed up to spend their travelling time in different ways. For example, passengers could work on their laptops, catch up with family and friends over the phone, read a book or watch a film – the possibilities are endless. We might even start to see travel time counting towards working hours, as employees are able to do tasks on the move and get to the office later. However, as with everything, there will also be mitigating factors the automotive industry will need to consider. For instance, the impact on the insurance industry and the ripple effect to other industries, including hotels.

The shift to AI represents a transformative period in the car industry

Manufacturers’ business models

 

Megatrends Q1 2018 - Capgemini AI senses

It is not just cars that will be transformed by AI; manufacturers’ processes, sales and business models will change too. Unlike industries which are faltering from the pressures of tech giants and start-ups taking market share because of AI-related innovation, automotive incumbents retain a distinct advantage. The main reason for this is the automotive industry’s high cost barrier to entry; any new entrant would have to invest significant capital in large-scale research and manufacturing facilities before being able to produce their first car. And even then, producing vehicles at scale remains a challenge: Tesla, for example, has only managed to produce a very limited number of cars each quarter up to now.

In manufacturing plants today, there are thousands of robots and cobots (collaborative robots), all with functions important to the car making process. Each of these collects data which AI and machine learning is able to use to predict failures and enable engineers to pre-empt downtime, which improves uptime and productivity. The same is true of when driverless cars go on the road. Autonomous vehicles also collect immense amounts of data, enabling engineers to advise when cars are likely to break down and fix faults before they occur.

While humans get used to the concept of not having to sit behind a steering wheel again, car manufacturers have a great opportunity to grow their businesses on the back of this new technology

AI will also enhance manufacturers’ sales and marketing processes. Manufacturers are sitting on a huge amount of information from cars already on the road, but are unable to harness it for sales purposes with insights buried amongst mountains of data. AI will change this. For example, with the ability to sift through the huge amount of data that their cars produce, AI will enable marketers to identify which non-vital functions are most useful to owners and target their campaigns as such. Furthermore, AI will be able to distinguish the most valuable sales leads based on past behaviour and propensity to buy, meaning traders can direct their efforts towards customers who are most likely to make a purchase.

Regulation and legislation

Advances in AI technology will also spell change in the way the automotive industry is regulated. Driverless cars will require legislation updates in order to come onto roads, and across areas such as driving licences, insurance and the rules of the road. For example, while adoption is still in progress, driverless car lanes will be necessary to ensure the safety of all vehicles. There will need to be rules put in place around driverless vehicles and junctions- no one wants to be stuck behind large trucks at the best of times, but a convoy of eight driverless trucks could block junctions and stop cars from leaving roads at the right point. More seriously, tough decisions need to be taken by regulators on how AI should be programmed in the case of traffic incidents;  when collision cannot be avoided, which car (or person) should come off worse?

Within the next 20 years it will take over from humans as the main entity behind the wheel. However, AI in the automotive industry is more than the technology merely passing its driving test

Insurance policies, too, will have to change. While car insurance is mandatory now, this could change to ‘passenger insurance’, whereby every person who enters a vehicle needs to be insured for any injury caused to themselves while in an autonomous car. In addition, premiums should go down as autonomous cars make roads safer and less unpredictable.

Finally, while AI learns to drive, the need for humans to pass their tests will diminish. To the joy of teenagers around the world, there will be no need to take a rigorous driving test with the availability of autonomous cars. However, that’s not to say there won’t be a requirement to have some kind of instruction on how to use driverless cars and what to do in the case of emergencies. This will be a huge regulatory debate within the coming years as autonomous vehicles become omnipresent.

The shift to AI represents a transformative period in the car industry. While humans get used to the concept of not having to sit behind a steering wheel again, car manufacturers have a great opportunity to grow their businesses on the back of this new technology. From providing the autonomous vehicles themselves to improving manufacturing processes and working with governments to update legislation, the manufacturing industry is entering a new era of opportunity.

 

Welcome back , to continue browsing the site, please click here