Hardly a day goes by without some sort of announcement about another application for autonomous vehicles. Just the other day I saw a commercial about a self-driving beer truck. It still had a driver, but he was puttering around in the cab as the truck rolled down the highway. That is great for the concept, but highly embellished. It is a great example of what the possibilities are, however. Today, as I was polishing up this missive I got a feed from TU Automotive with a headline that stated, “Have car giants admitted driverless tech too big to take on alone or too quickly?” Still another vector from the design tech sector noted “Autonomous Cars Face a Slew of Sensory, AI, and Communications Challenges – getting driver-less vehicles to achieve their full potential will be no small undertaking.”
I write a lot about autonomous vehicles (AVs). I find them fascinating, especially their existing and potential connection to the wireless ecosystem. I also know how this stuff works, and if you follow this at all, you know that we are a long way away from truly steering- and pedal-less vehicle. Of late, there are signs that the Kool aid-drinking window is finally shrinking up and the reality of what it takes to make this happen is settling in.
There are still many challenges – some we haven’t even addressed yet. Those we are working on, like object-detection, communication, navigation, safety and security are still in their infancy. Before these become the driving (no pun intended) force in AVs, they will have to make some leaps in advanced hardware, and integrate with advanced AI.
Today, AVs, outside of the lab or beta tests, are just driver-assist platforms. Some further advanced than others, but driver assist, nonetheless. Production vehicles have various iterations of platforms such as adaptive front lighting, lane departure warning, adaptive cruise control, self-parking, blind-spot detection, and emergency-assist braking. That is about it. Moreover, none of these use a lot of AI – and that is the key. The next generation of AVs will require much more robust systems, such as global navigation satellite system (GNSS), and advanced sensing capabilities. And that still won’t put us in the fully autonomous game.
The problem comes with one-way communications. Things that go one way, such as GNSS are critical elements in the AV wheelhouse. But just knowing where vehicles are, both from satellites and terrestrially, isn’t enough. The satellite may be able to pinpoint your position and the sensors tell you what is around you, but they are not going to tell you what is going on 10 cars ahead, without some sort of intercommunication among vehicle. Even advanced sensors, with advanced AI, while they up the ante an order or magnitude, still cannot integrate with the infrastructure, and respond to the fallout of non-linear, split-second incidences when they occur. That requires feedback.
However, there is a lot of excitement in this segment. What is under the advanced driver assistance systems (ADASs) currently on the market are ramping up in technology. It is great progress. And for the next few years, such systems’ components will be refined to make them more reliable and less expensive, so the reliability factor goes up and redundancy becomes more affordable.
One example comes from Bosch, who claims they have radar that can tell the positon of the vehicle within centimeters, within its lane. Bosch calls this their Radar Road Signature platform and it uses some AI along with complex processing algorithms that are cloud based. The vehicle only captures the data.
A red flag pops up here in that if there is a corona mass ejection, for example, and the wireless infrastructure takes a hit, there better be a warning system that alerts the driver their ADAS is offline.
This is one example of some of the complexities that are involved in AVs. There are others but we will save them for another day.
Ernest Worthman is the Executive Editor of Applied Wireless Technology magazine. A Life Member of the IEEE, his 20-plus years of editorial experience includes being the Editorial Director of Wireless Design and Development and Fiber Optic Technology, the Editor of RF Design, the Technical Editor of Communications Magazine, Cellular Business, Global Communications and a Contributing Technical Editor to Mobile Radio Technology, Satellite Communications, as well as computer-related periodicals such as Windows NT. His technical writing practice client list includes RF Industries, GLOBALFOUNDRIES, Agilent Technologies, Advanced Linear Devices, Ceitec, SA, and others.