Opening The Doors to the Sustainable Connected Car
24 April 2016, by Joel Hoffmann, Open MCB Community – At the recent Automotive Megatrends event in Detroit, John Maddox of the University of Michigan Global Technology Center insisted that communication is the key to solving the congestion problems and the safety issues that would eventually result from self-driving cars. We’ve been focusing a lot on how to get to those self-driving cars. But without communication, he indicated that potentially these cars could be so independently controlled, perhaps by computers, that they really wouldn’t be that much safer than human drivers. Particularly, they may not contribute to solving congestion problems because you won’t have control over them as a community. Independently there will not be an optimization of the system, but rather an attempt to optimize the experience of individual cars. Mr. Maddox should know since he used to be associate director for NHTSA, and worked almost 20 years for Ford and VW. Under his leadership, UofM is running the largest simulated city to test autonomous cars in real world conditions, including Midwestern snow and Michigan potholes.
You can’t collaborate amongst cars on the road unless you have communication. Communication is essential, even for just locating where you are, and what else is around you. Communication is much more than connectivity.
We need all three of those things. And yet, the early attempts to introduce communications to the car haven’t been very successful. Most of the systems that were introduced have been limited to one brand of car, or one specific model or line of cars, and there’s never really been an effort to pull all the potential billions of cars in the world to communicate with each other through closed telematic systems.
So we’re still seeking the connected car in a universal sense – with doors wide open. And a connected car is certainly going to be a big part of the autonomous car. How are we going to get there? Collaboration occurs at three levels. At the industry level, which is the leadership role that the automotive industry needs to play. At the community level, which is actually much larger than the industry level and will certainly include independent software developers. And finally, at the individual level where buyers of the connected and autonomous cars will pay the premium price demanded. I’ve seen this kind of collaboration happen in my work within alliances, where open-source or open discussions take place amongst competitors.
Of course, we want smart cars to communicate with smart people and smart things. The Internet of things is promoting a communication system for everything we own, and collaboration is needed across that. The wireless industry, which is very involved with phones and devices have had similar challenges in the past and the automotive industry can leverage that.
Let me introduce the auto industry to the Open Mobile Alliance. It was started in 2002, so it’s been on the road for a little while, with a few well earned battle scars to prove it. The wireless industry members sit together in deep technical discussion about how to standardize commodity parts of the industry. This is not a new concept to automotive. It’s done everyday. In the steering systems, in the braking systems, in the tires. Everything including the wrench dimensions in your garage toolkit to thread spacing of bolts sold at Home Depot are defined by automotive standards. Lots of standards have been introduced by automotive, but they haven’t blended successfully with the work of the wireless industry.
These wireless companies want to do business with the automotive industry. Some of them have been successful, but mostly on a case-by-case basis. They need to have an umbrella relationship to avoid not only a market challenge but a safety risk.
An exaggerated version of that risk is something we’ve seen a ton of in the news – hacking cars and breaking through security. That’s been attempted because of the immense complexity of the car and it’s connectivity. The hackers aren’t at fault for this, they’re just finding holes that were left there because there wasn’t enough collaboration amongst the software developers and testers.
I’m sure you’re all familiar with the well-reported car hack that was published in Wired magazine. Some of the companies in the Detroit area don’t want to hear about it anymore. But it establishes that if you’re given enough time, you can hack into pretty much any car. The interesting thing is that if you read the text there are three things you need. First, you need to have a car. You have to have the exact same car you are planning to hack, and you have to work on it for about a year, to reverse engineer not only the software but the silicon. It’s a very difficult process, but you get through it, figure it out, and after that you’re able to compromise the car’s network. Then you can do some pretty interesting things like sending messages, stopping and starting the car, changing the radio stations, or crashing it remotely into a less connected car.
I think this has changed the Internet of things into the Internet of targets, and people are afraid. They’re worried about it. How do we fix it? More recently Wired.co.uk published a lesser-known article highlighting the Nissan leaf cloud hack. This one is interesting because they did not require access to the car, ever. All the hacker had to do was know was a VIN from an Nissan Leaf and use a browser with access to the public internet. This hack was performed across the world on different continents. And it was possible because Nissan had not embedded critical security into their app, their server, or the car.
Now, was that the carmaker’s fault? I’m sure many people would think so. Was it dangerous? Well, the worst thing you can do to that Leaf is to remotely turn the AC up while parked, on a sweltering day in Arizona, while also heating the seats, and run down the battery before the driver comes back so that now he’s got an undrivable car. That’s about the most damage you could do. The app didn’t control the car. But what if the app did control the car? What would happen? Much more. How could Nissan have improved that? They could have implemented well proven protocols that were developed over many years of other people testing them. Also, if they planned for over-the-air updates they could have fixed it before the article was published. Very few car companies have this figured out yet.
The wireless industry depend on standards development organizations that heavily work together to figure these things out. There are different ways to do it, some work within closed organizations where you have to own access to these standards. Others such as Open Mobile Alliance, known as OMA, make their standards publicly available. With all the related standards organizations with catchy 3-5 letter names that OMA collaborates with, they are somewhat of a “one-stop-shop”.
One of the most successful standards implemented in wireless is called OMA DM or device management. OMA DM is used in two million devices across the globe including every single phone in your pocket, every tablet. It’s critical to the provisioning and remote updating of the OS and apps in wireless devices the whole world depends on. A derivative of that is called lightweight M2M, which is designed specifically for the Internet of things.
So why can’t we get the brainpower of this wireless community that’s figured these things out to work on cars? I think we can. That’s something we set out as a goal. And we, OMA, are creating an industry incubator called OMAuto. We want to incubate the collaboration between the automotive industry and the wireless industry. And we’re inviting automotive industry companies, individuals, and experts to come into this incubator. It’s going to be a series of technical meetings, mostly on the phone, sometimes face-to-face. The duration of the incubator will be six months, and the outcome will be a set of recommendations that will go to the OMA standards working groups. It’ll be complicated if we try to tackle all the problems at once, so the members of the incubator will decide how much they can accomplish in six months. And if it works, we’ll repeat the process. Nobody’s going to spend any money. Nobody has to join any clubs. You don’t have to pay to be a member; you don’t have to join an alliance. You do have an application to complete but it’s relatively lightweight, see the new landing page at: http://openmobilealliance.org/about-oma/work-program/omauto/.
And with that you have the ability to influence that future of safer and more connected cars. Everyone has a chance to make a difference in your driving; or non-driving future.
To get involved, please contact me at [email protected] or meet me in Paris at the GENIVI All-Member Meeting April 26-28, 2016.