Thinking cars and non-thinking humans
By definition, artificial intelligence is when machines are increasingly able to think, and more importantly, learn the way humans do.
In the automotive world, there are very clear examples of where this can be useful. The easy answer is in automated or autonomous driving, where the system can learn and react based on different inputs (road, traffic, demand).
We are already seeing similar systems in place in certain uses such as transmissions that “learn” how we like to drive and adjust accordingly.
Subaru’s EyeSight is an example of how a simple system such as cruise control (we think it’s simple now, but it was once ground-breaking) can be enhanced by using sensors and cameras to gauge traffic, speed, proximity, closing velocities, and so on.
The different lane-following software already in place in certain cars is a base from which further developments can come.
And the various autonomous driving systems themselves that we have tested are eye-opening.
They can be both computer-like in their precision, but also human-like in the way they are actually used.
The Japanese systems, for example, tend to be very community-oriented, whereas the American systems seem to focus almost completely on the individual (or the individual car) looking outwards.
But the effects can be deeper and more far-reaching than we may even realize without necessarily affecting our drive itself.
German car maker Audi recently announced that they will apply what they are calling “machine learning” to the processes within their machine shop.
Quality inspections for marks and cracks on sheet metal will be done using this enhanced system, which will allow more quick and precise identification of issues.
Audi is looking forward as increased use of complex shapes and forms (see its sister company Bentley and the sharp edges of its new Continental GT) will make this area more demanding of precision and consistency.
Whereas current systems do this by matching captured images, the software being put into use will be based on a system with more unstructured and high-dimensional amounts of data.
Artificial intelligence can affect the industry in other ways as well, systematizing processes in areas such as finance, planning, logistics and the like.
Predicting when parts and supplies need to be available would be a welcome improvement in the eyes of many consumers, but also to those of the dealerships that need to stock pieces that may not move right away.
Everything could be more fluid. Heavy number-crunching departments such as financing and insurance could benefit heavily from more accurate forecasting that could as well trickle down to the wallets of all of us.
This doesn’t mean that things will be cookie-cutter all over the place. There is room for individualization, and the nod to the proper heritage and DNA.
I personally feel that the early lane-following systems of Porsche were much more in tune with me than were those of other German brands. But that’s all a question of how you program.
Having said all that, the increased computerization and complication of our vehicles and our lives is very double-edged.
We bemoan the lack of simplicity of maintenance and repair, but we love the increased safety, comfort and flexibility.
What may seem to happen is that we as a community will end up with certain parameters in which we must work or which we must accept (more complicated repair, less places that can do the work, more environmental friendly, more safe travel), and it is within those spaces that the car companies will work to fine-tune their products to appeal to us on a more individual and even passionate basis.