It’s pretty difficult to miss the hype and speculation around autonomous cars these days. The idea of our roads full of self-driving cars may still seem like science fiction, but as a growing number of automotive giants assure us, this reality may be closer than any of us imagine.
In August, Toyota and Uber became two of the latest industry players to join forces on an autonomous vehicle partnership. Toyota is to invest £387 million to help Uber develop a fleet of driverless cars for its ridesharing business, in a move being seen to help both companies boost their profiles in the autonomous vehicle sector.
According to those in the know, global investment in autonomous cars is to hit $557 billion by 2026, as every automotive and tech company from Audi to Apple, Tesla to Samsung, are working on innovations they hope will shape the market.
So, how much do we know about how autonomous cars will work? Let’s get under the bonnet of the automotive industry’s high-tech future.
To give them their own definition, ‘autonomous’ vehicles – or those with the freedom to act independently – must be able to get to a destination and park without any human intervention, and without hitting anything.
It’s artificial intelligence (AI) that makes this possible; a system capable of sensing and interacting with its environment, reacting to and, most importantly, pre-empting changes to its surroundings. In short, the AI system has to act just like a human driver would or, more to the point, better. A fundamental prerequisite of autonomous cars as a concept is that they must make our roads safer, reducing the risks associated with human error and misconduct.
AI applications ‘learn’ through exposure to data, vast amounts of it. Through this deep learning technique, vehicle technology can build on its experience of the data it has processed and modify the way it reacts accordingly. Car manufacturers and tech companies are busy putting the AI systems inside their prototype autonomous cars through their paces, presenting them with every imaginable obstacle and driving scenario so that they can ‘learn’ how to deal with them. In this instance, the ‘data’ enabling AI awareness is anything that can lead to a collision and obviously, until these systems are as infallible as they can be, no fully autonomous car will make it on to our roads.
Having said that, self-driving pioneers Waymo are said to be close to launching the world’s first fully autonomous taxi service in the US. Their fleet of autonomous minivans are already offering free rides to a test group in Phoenix, Arizona, albeit with human ‘back up’ drivers that monitor every journey, and there are plans to extend to the general public soon too. The minivans are said to be logging 25,000 miles a day on public roads, creating invaluable data that enables them to get smarter all the time.
An incredibly complex combination of data-led AI systems have to work together seamlessly to make autonomous driving possible. As we’ve mentioned, autonomous car technology must be able to not only react to events around it, but also predict them before they happen and take evasive action.
This is only one type of data autonomous vehicles rely on though; they also need access to up-to-the-minute mapping data in order to navigate the most efficient route from A to B, and to recognise and interpret road signs and changing light conditions. Most autonomous car companies aim to achieve this through three interconnected sensory systems; video cameras, radar and lidar, the last of which uses light waves to read the physical environment and create a continuously-changing 3D map.
What’s more, if, as many are predicting, one in four cars on the road are self-driving by 2030, they will need to be able to communicate with each other and share real-time data. Vehicles with ‘smart’ capabilities – those connected to the Internet of Things (IoT) – are already emerging on to the market, so it follows that autonomous cars will use this technology to create an intelligent multiple vehicle network.
Taking this idea and running with it, Volvo has launched its vision for a fully electric and autonomous fleet of truck cabs, which would use a cloud-based system and a central data hub to control the cabs en-masse. In the same way, it’s likely that if and when autonomous cars truly take off, they will offer a lift share service rather than a product. Maybe we all won’t own autonomous cars in the near future, but we may be able to order one as easily as we do an Uber right now.
None of what autonomous cars should be able to do, would be possible without data. The more data prototype vehicles are fed, in the form of the unpredictability of the road, the more they learn and the more data they generate that can be used to refine autonomous vehicle technology further still.
Consumer data can be used to inform marketing campaigns in a similar way. With the right consumer data insight behind them, marketers can better identify and target those that represent the most commercial opportunity, in ways that offer those consumers something of value. The data these campaigns produce can then be used to constantly improve and hone marketing activity moving forward, consistently providing a quantifiable business intelligence edge.
At Quant, we have the data insight capabilities to drive your business towards better customer relations and marketing ROI. Don’t hesitate to get in touch if we can help put you back in the driving seat of your data marketing.