The phrase "car modification" bring to mind images of grease-stained burly persons clad in oil-spotted coveralls, toiling away at the aluminum and steel heart of some piece of American muscle. Those who coax more out of a vehicle are seen as artisans. Their canvas is an engine block. Their studio, a garage.
Zac Nelson doesn't quite fit that stereotype. He may be the son of an ironworker who grew up in a home full of engines in various states of assembly, but now, as a research engineer at Ford, Nelson works in a pristine fluorescent-white office, wears neatly-pressed slacks and shirts, and manipulates lines of code instead of valve timing. While Nelson spent his early years at MIT fabricating and welding the chassis for a Forumla SAE car, his senior project focused on software—designing algorithms and microcontrollers for an adjustable headrest.
"I'm pretty familiar with the metallurgy and mechanics of materials, but the electrical side of engineering is a new frontier for me," says Nelson, who graduated from MIT this past September.
In this respect, Nelson is part of a new class of automotive mechanics that seek to improve the automobile with data sets and text editors instead of wrenches and milling machines. "Look around any high school. For every person who modifies an engine there are about five to ten people who are doing complex coding on a laptop," says K. Venkatesh Prasad, group and senior technical leader of vehicle design and infotronics within Ford's Research and Innovation department.
In his first major project for Ford, Nelson zeroed in on the shift knob. "Historically speaking, [the shift knob] has just been a piece of plastic," he tells PopMech. "Maybe there's some metal weights in it for feel and performance, but other than that there's no electronics—it's not really considered a smart device."
Using a 3D printer, two vibration motors scavenged from an Xbox 360 control, and the Ford-developed OpenXC platform—software that can translate signals generated by a car's controller area network (CAN) into usable programming code, much like an API for autos—Nelson created a shift knob that vibrates at the ideal moment to change gears. And unlike the aftermarket shift lights commonly found in racing cockpits, Nelson's "haptic shift knob" isn't tied to a static RPM. Connected to the shifter is tablet running an Android app that constantly compares engine speed and accelerator position to the car's torque curve. Hit the gas hard and the knob will wait until later in the RPM range to signal a gear change. Go easy and it'll find a shift point that's more economical. In a way, while Nelson's is tinkering with the car, he's also altering driver behavior.
Released to the public at the Consumer Electronics Show this past January, OpenXC is Ford's way of luring Silicon Valley into the automotive sphere. "There's a lot of people who can write apps," says Prasad, who left a research position at the Menlo Park, Calif.-based RICOH Innovations for Ford in 1996. "What OpenXC does is open up cars to people who want to do pure software work."
Fittingly, Prasad has Silicon Valley-sized ambitions for the software platform. One example he gives: an app that aggregates wiper blade usage pulled from OpenXC to provide accurate localized weather reports that even the National Weather Service cannot rival. "The fact that you have a wiper blade that's turned on is really valuable information, especially if you can aggregate a four or five of those data points a mile or so up the road," he says.
The catch is getting those four or five data points. While a user can download the OpenXC software and tinker with the API for free, harnessing the power of your car's data (and in Prasad's example, the data of other cars) requires a $110 vehicle interface (VI), a piece of hardware that reads CAN signals and translates them into values you can plug into a program. Even more restrictive is the VI's firmware, which is only compatible with Ford's CAN (CANbus signals differ from manufacturer to manufacturer).
According to Prasad, other manufacturers haven't joined forces with Ford on OpenXC, despite it's open-source nature. He cooly encourages them to. "Everyone is welcome, and it only gets stronger when other manufacturers get involved," he says. "I can't speak on their behalf, but it's up to them. They're more than welcome to take a look."
Nelson's haptic shifter is a good example of what can be accomplished by exploiting vehicle data, but the only on the OpenXC website are from Ford employees or interns. It remains to be seen what will arise organically from OpenXC. "The future is the future," Prasad says. "So who knows."