As reported on Wired.
BY JOSEPH FLAHERTY
- Automatic has developed a new quantified-self gizmo that promises to save drivers thousands of dollars per year and stall the greenhouse effect by reeducating drivers in real-time. Photo: Automatic
Do you know how much money it costs you to slam on your brakes? Or the annual cost you incur by speeding? Probably not, and even though the average American drives 13,476 miles per year, aside from a brief driver’s ed course in high school, few get any formal training. The result is a lifetime of bad driving habits and wasted fuel. Fortunately, a San Francisco-based startup called Automatic has developed a new quantified-self gizmo that promises to save money, and stall the greenhouse effect, by reeducating drivers in real-time.
The $99.95 dongle plugs unobtrusively into your car’s OBD-II port, pairs with a companion iPhone/Android app via Bluetooth, and monitors three key behaviors—sudden starts, speeds over 70 miles per hour, and hard stops.
Monitoring these three behaviors isn’t the full picture of someone’s driving style, but even small improvements in these basic metrics can be impactful. “An average driver can save 30% by driving more smoothly,” says Automatic co-founder and product lead Ljuba Miljkovic. “Which is nuts if you think that the average driver spends $3,000 on gas annually.”
Retraining drivers requires real-time behavior change, but presenting feedback visually at 75 mph is difficult, not to mention illegal in states with distracted-driver laws. Instead of forcing users to mount their phones like a GPS unit, Automatic developed an interface featuring three crisp tones that alert drivers to their transgressions while making a commute feel like a game of Mario Kart.
Automatic is a handy tool for gearheads, but was heavily influenced by egghead Stanford professor B.J. Fogg. His Behavior Model theory states that effective behavior change requires motivation, ability, and triggers to come together simultaneously.
Automatic creates motivation by pointing out how bad habits waste money and provides audio triggers that condition the driver to take appropriate actions. The results are impressive—the software was designed to give an average driver a score of 60 out of 100, but Miljkovic says observant drivers quickly reach the low 90s.
While most app designers obsess about creating seamless flows and and spurring downloads, the Automatic team is trying to do something much more ambitious—redesigning human behavior using the app as their tool. This requires a delicate balance of technology and psychology. The interface has to work seamlessly, but also educate drivers about their bad habits without shaming them. “Every one thinks they are a good driver,” says Miljkovic. “But generally, the more aggressive someone drives the more likely they think they’re a good driver.”
Automatic avoids value judgements by making everything from the naming of behaviors to the sound design factual and empirical. The app makes the monetary cost of habits understandable and actionable, but leaves the decision about what to do firmly in the driver’s seat.
The chirping clamshell provides real-time feedback, but the app acts like a longitudinal dashboard between trips. For instance, an Automatic engineer used the onscreen information to calculate that by dropping his average speed by five miles per hour would add five minutes to his commute, but save $50 a month on gas. For those trying to live on a tight budget, the app becomes an offline tachometer.
The app makes the monetary cost of habits understandable and actionable, but leaves the decision about what to do firmly in the driver’s seat. Photo: Automatic
Beyond it’s ability to fill your wallet, the app is chock full of small touches. Setting up the dongle can be a pain since the OBD-II port is usually hidden in a dark, hard-to-reach spot, so the setup interface has a button that turns on the phone’s flashlight.
When early users were confused by the ordering of past trips on a timeline, Miljkovic added a little car with headlights that come on when a user moves forward in the history, creating a link to the real world. “We tend not to do much in the way of decoration, but we like small touches that help reinforce what the product is about,” says Miljkovic.
Automatic also has put great thought into features you’ll never want to use. If your car is in an accident, the system will text a family member with your location and offer you live assistance on the phone. A “Check Engine” light translator lets you know if scheduled maintenance is required or if your car has blown a Johnson Rod. But aside from emergencies, the app is deliberately anti-social to ensure the device can’t become a beacon for stalkers.
The future of connected cars is bright, and if Google has its way, we may soon be ferried around in algorithmic harmony by self-driving cars. Until then, Miljkovic is happy to level the playing field between the Tesla crowd and those of us driving more modest models. “I’m thrilled that people in a 1998 Honda Accord will have features equal to those in a high-end Mercedes.”