Do Self-driving Cars Hold the Key to a Widespread IoT?

In 2014, Continental Tires developed tires that talk to you . The innovation, dubbed eTIS (electronic Tire Information System), consists of sensors embedded beneath the tire tread. The sensors relay information about when your tires are underinflated, when tread is too low, and when your car has too much weight in it from a heavy load. This new entry in the annals of IoT tech was relatively quiet and unglamorous. Yet, it forecasted what we’re seeing now. Car manufacturers and tire manufacturers are throwing millions of dollars into technology that will enable a widespread internet of things.

Call it necessity facilitating innovation; as I reported in an earlier post here, 1.2 million people die in auto-related accidents every year. That means safety is in high demand. One way to increase safety is to embed things like tires with sensors that can communicate data with a car’s onboard computer. Another way is to replace humans with AI to create self-driving cars, which will hopefully do a better job than we do at driving.

For self-driving cars to truly succeed by 2020, the IoT needs 4.5 million developers. That’s because a comprehensive IoT infrastructure ”in which smart cities talk to smart cars ”will help driverless vehicles navigate through the infinite variables on the road. Auto manufacturers and other interested parties ”Uber and Waymo, to name a few ”are doing a great deal of legwork to kickstart a widespread IoT. I predict that the auto world’s push to advance new technology will be the primary reason we’ll see the IoT take off.

It’s not just Uber and Waymo racing to make this happen. Here’s a look at developments that could see us living in smart cities soon.

Meet Olli

In 2016, a company called Local Motors unveiled an electric self-driving minibus called Olli. Complete with an app, Olli is an on-demand autonomous vehicle that Local Motors manufactured with 3D printers in a micro-factory. The vehicle’s interface is IBM‘s Watson, an AI supercomputer.

In 2017, Local Motors plans to launch Olli in Las Vegas and Miami. IBM designed the interface to talk to people. Passengers can ask it questions, and it provides information about their destination’s attractions. The vehicle employs an array of sensors, including lidar, and a haptic sensor that makes it automatically stop upon contact with anything in front. The vocal interface uses cognitive computing to adjust to passenger input, and it will eventually offer advanced personalization through passengers’ social media profiles.

Local Motors’ CEO Jay Rogers plans to have 20 micro-factories running by 2020, with the capability of producing 60,000 units a year. The first Olli took two weeks to produce through 3-D printing.

Cars That Talk to Each Other

The World Economic Forum reports several developments that will influence both self-driving cars and the IoT. First, the US Department of Transportation (USDoT) wants to require artificial intelligences in self-driving cars to talk to each other. In that sense, the many autonomous vehicles in the works from the likes of Ford, GM, Google, Uber, and Tesla would have to become an interactive IoT, with the ability to send data to each other, despite the fact that they’re from competing companies.

Secondly, Sam Abuelsamid, senior research analyst at Navigant Research, says that self-driving cars won’t just be able to rely on data from their own visual sensors; they’ll need to rely on maps as well. Visual sensors can be disrupted and rendered blind by bad conditions. That’s why mapping experts from TomTom are working with AI developers at Nvidia on extremely granular, highly-detailed maps.

<iframe width=”560″ height=”315″ src=”https://www.youtube.com/embed/86U4sxCpRf8″ frameborder=”0″ allowfullscreen></iframe>

HD maps will integrate vehicle lidar with AI software to create real-time road-analysis and decision-making. Yet, what you’re seeing here from TomTom is merely the tip of the iceberg, because data input is only coming from the car’s sensors and from TomTom’s real-time map overlay:    

Source: TomTom

What’s missing from this picture? Data from other self-driving vehicles, which the USDoT is planning on incorporating. Although it’s vague, TomTom’s sensor input may cover data from external sensors wherever TomTom can get them. As all the breakthroughs in self-driving tech surface by 2030, it will only make sense for the USDoT to embed sensors in stationary objects to communicate with vehicles. This isn’t a matter of luxury and convenience, like a smart home ”it’s a matter of life and death for human beings.

To avoid the primary problems for self-driving cars ”including impulsivity of humans and variability of conditions ”there will have to be a concerted effort to mobilize the IoT, and it will have to happen fast. That’s why big data jobs and deep learning jobs are popping up in abundance, and they are centered on AI and the IoT.