Audi unveiled its ideas for streamlining automotive
transportation at the Los Angeles auto show when it introduced new technologies
aimed at improving drivers’ information and decision-making capabilities.
The Audi Urban Intelligent Assist project is a three-year joint effort between Audi AG, the Volkswagen Group Electronics Research Laboratory (ERL), in Belmont, CA, the University of Southern California, the University of California at Berkeley, and the University of California at San Diego. The University of Michigan’s Transportation Research Institute also participated initially but has left the project.
The project, which wraps up at the end of 2013, aims to find technologies that can help reduce traffic congestion, improve safety, and reduce stress on drivers, according to Mario Tippelhoffer, Audi’s lead engineer.
Development efforts fall in two primary areas: the Audi Driver Centric Urban Navigation application and the Audi Urban Assistance app.
“We are developing smart phone apps that help you plan your
trips before you get into your vehicle,” Tippelhofer explained. The USC team collects data from multiple
sources to provide predictive information for drivers using the Audi Driver
Centric Urban Navigation application suite.
The Time-2-Start app runs on drivers’ mobile devices and lets them know how long to expect it to take them to make a planned drive so that they know when to start. It uses real-time traffic data along with historical traffic information for the location and time of day.
The Smart Parking app recognizes drivers’ parking preferences and habits and searches for available spots in the destination area, automatically providing routing information to the space.
Predictive Traffic forecasts traffic conditions using current traffic information, historical data, weather, and event information. That means that when the home team has a playoff game, drivers are warned to avoid trying to drive through that part of town.
The Predictive Destination app guesses drivers’ intended destination based on their history, location, and time of day to provide information about the route the app expects the driver will take. “Most people go to places on a regular basis,” said Tippelhofer. “That is a really repetitive pattern we see. Most of the time you go to only five different locations.”
Naturalistic Guidance recognizes that having a robotic voice say to prepare to turn left in 400 meters isn’t terrifically helpful to many drivers, but a nav system telling them to turn left after the 7-11 is much better, so that’s what naturalistic guidance aims to provide.
“A person giving directions would say, “Take a right after the Starbucks,’” Tippelhofer explained. “We want to use prominent landmarks like churches and restaurants to make route guidance more natural.”
The Seamless Navigation app addresses the fact that we can sometimes reach our parking space for a destination but still need some help finding the actual place, so it gives drivers walking guidance all the way to the door.
The other category of applications, Audi Urban Assistance, looks at scanning around the car to provide information about tactical driving conditions. USC, UC San Diego, and UC Berkeley teams are contributing to these applications.
Merge Assist is a new application designed to help drivers get the speed and timing needed to merge smoothly with surrounding traffic, giving them a target speed on the instrument panel and green LEDs in the side mirror to indicate it is time to merge. It acquires the necessary information about surrounding vehicles using a combination of video cameras and radar, said Tippelhofer. “We have some really sophisticated sensors that can monitor 360 degrees around the vehicle,” he said.
Lane Change Assist is a more advanced version of today’s blind-spot warning technology. It improves on current products by watching for fast-moving vehicles approaching in the adjoining lane and not just vehicles currently in the car’s blind spot.
Attention Guard attacks the problem of driver distraction, using cameras to watch the driver and alerts to regain their attention if they are not focused on driving. “We want to keep you safe if you are not paying attention,” said Tippelhofer. “We want to bring you back to your driving task.”
How to do that is under discussion. “We are looking at [human-machine interface] solutions for different ways of telling drivers.” In a worst-case scenario, the car could even determine that the driver is not going to respond to a crises and intervene automatically, he added.
Each of these applications is on its own schedule, depending not only on the state of the technology but also on the available support infrastructure, where that is relevant. “Some of the applications we are developing could be applied rather soon to an automotive environment,” he said. “Some of them will require some infrastructure,” such as traffic and parking sensors.