Artificial intelligence will be part of automobile's future

By February 29, 2016
Une voiture connectée du constructeur Tesla Motors.

Given the increasing range of functionality available in connected cars, it is arguable that we now need a virtual assistant (VA) as a co-pilot. In the longer term, artificial intelligence (AI) may well be the key to allowing fully autonomous cars on to our roads.

A great deal has been written recently about self-driving cars, but their future is still very uncertain. When will they be allowed to travel freely on our roads, apart from in the test phase? When are we going to see widespread consumer adoption – 2020, 2030 or 2050? In any case the transition is likely to take place in stages – three phases is the number commonly mentioned – beginning with the advent of the connected car, which is already with us today. Connected cars are able to communicate with the driver, provide updates from the various car indicators such as oil and coolant levels and brake pad wear and information on weather conditions, and actually help the driver to drive the vehicle, at the same time enabling him/her to go on to the Internet to obtain the latest news, check Twitter feeds or launch a playlist on Spotify. However, all these novel options call for a certain level of attention from the driver and risk distracting him/her from the road. So how can the range of services offered by the connected car be reconciled with road safety? ‟A connected car comes ready-made with a set of distractions, so a virtual assistant is especially necessary in this area,” argues Ron DiCarlantonio, CEO of iNAGO, a pioneer in the field of AI systems.

Controlling your vehicle intuitively

A VA can provide access to a range of functions available in the connected car through voice commands, which allows the driver to keep his/her eyes firmly on the road. Moreover, ‟the virtual assistant allows you to control your vehicle completely naturally and intuitively, using a mixture of voice commands and gestures. It can also interpret the driver’s facial expressions,” points out Ron DiCarlantonio, adding: ‟In addition to having total mastery of the vehicle, the virtual assistant enables you to receive the right information at the right time, providing both content and services.” The VA can for instance look up an address for the driver, find a place to park in the neighbourhood and remind him/her of what s/he has arranged to do that day. Starting out in Ontario, Canada, iNAGO moved to Japan in the early 2000s and has already created AI systems offering various different types of services, including virtual telephone operators for call centres and customer service centres, virtual teachers to teach English to children, and, more recently, VAs for drivers. Having worked for some time in this field with Japanese automobile manufacturers, the company launched its automotive AI products on to the North American market last year.

AI and autonomous vehicles

Artificial intelligence is set to become an essential element in the self-driving cars of the future. ‟Autonomous vehicles will use ‘deep learning’ to process huge quantities of data, to determine what types of situations generally lead to an accident, and react  accordingly when that kind of situation arises,” predicts Ron DiCarlantonio. ‟Computers have a number of advantages over human beings, which make them better drivers: they have more sensors, they see better, they aren’t emotional, they never get distracted… But we’re much better than they are when confronted with a situation for the first time. So if vehicles are to become totally autonomous, computers will need to be able to react to the unknown and adopt the right approach when facing a situation they haven’t been programmed for.” This seems to mean building computers that are able to free themselves from our control and act independently… a scary scenario, perhaps?

Legal mentions © L’Atelier BNP Paribas