The Autoplug Project
Saturday, June 20th, 2009This post, about the Autoplug project Sunil and I are working on at the Real Time and Embedded Systems lab in the University of Pennsylvania, is long overdue. Its partially about current car control systems, partially about where we wish Car control systems were, and partially about the work Sunil and I have done. This content will also be on the Autoplug website that Sunil will be developing.
Until recently, cars had no integrated, universal system to monitor and control all of the various subsystems of the car. As of 1996, all cars manufactured must have an OBD II port, which provided a universal interface for owners and mechanics to extract important information from their cars. However, each car manufacturer used a different messaging protocol to communicate on the OBD port. So it was difficult to produce a system that was generalizable to multiple cars. Finally, it was stipulated that all cars produced after 2007 must communicate using the Controller Area Network standard.
Despite these advances in communication and information sharing, cars are still a black-box to most of the major stakeholders in the industry. Manufacturers have no idea how the car is being used. Owners don’t have a clue as to what is going on inside their car beyond what is displayed on the dashboard (if that…). And 3rd party parts vendors have no idea how to design interactions with the various parts of a car in advance of its release as the information for each model is different. Furthermore, this information is held secret and needs to be bought to be discovered. Not only that, but if an owner wants to add an upgrade to the car (a rear view proximity sensor), or program a new way to unlock their car (via their phone), they have to pay exorbitant costs to 3rd parties, and that is only if the changes they request are possible.
This is a travesty. How is that the rest of the technological world is innovating on the ideals of the PC(customization, extension, programmability) and the internet (information sharing, application sharing) while the automobile is stuck in the seventies?
That question can be best left answered by others, but suffice it to say that this is low hanging fruit that is the next logical step in the development of the automobile.
The next question to ask is, “Is developing the automobile to fulfill all of the ideals of the PC and internet worth it?”
We believe that the potential capabilities are pretty exciting. For example, let’s say your friend or family member just got their driver’s license, and will be driving in your car. You know your family member is not good at parallel parking, and you wanted to help th em out so that they don’t wreck your car when they drive into the city. So you buy a proximity sensor from Target, attach it to the front and rear bumpers, have it wirelessly interface with the rest of your car, and display the results on your lcd dashboard. Your worries of your little sister scratching your bumpers may not be completely gone, but they certainly are alleviated. That is how easy it should be to make changes to your car. If you want to control your car with your cellphone, there should be a free and open market that facilitates this change to your car. That’s why part of this vision that Professor Mangharam shared with us involves an application store, where 3rd party developers can create new devices or software to update your car. Like the IPhone commercials, there should be an app for that.
The difference is that with the automobile industry, there are already millions of dollars being spent on cars inefficiently and ineffectively. While the money would be pretty cool, that’s honestly for other peopl to think about. As Professor Mangharam said, “Let’s just make something cool.”
We’ll be releasing everything to open source so others can retrace the project, and I’ll post a link to the autoplug website as soon as its up.