Short biography
- 2001-2007: Computer science studies in Karlsruhe
- 2007-now: Research fellow at Research Center for Information Technology (FZI)
Working areas
- Cognitive vehicles
- Road sign recognition
- Traffic light recognition
Research interests
- Real-time image processing
- Machine learning
- Scene interpretation
Technical Cognitive Assistance Systems
| Environment Perception | Electric Vehicles |
| Advanced Driver Assistance Systems |
Autonomous Driving |
Segway Traffic Sign Mapping
- Mapping objects for OpenStreetMap can be rather tedious
- Let's automate it a bit
- Drive a Segway
- Recognize traffic signs automatically
- Assign an appropriate geo position
- Check for conflicts (already mapped)
- Upload to OpenStreetMap (user confirms)
Marble
- QML user interface
- Declarative Marble widget
- Custom layer
osmscout
- Road network
- Traffic signs
Towards Autonomous Driving
- Research vehicle: Audi Q5
- Additional sensors
- Video cameras
- Velodyne laser scanner
- Ibeo Lux laser scanner
- PMD cameras
- Oxford OXTS IMU
- Autonomous driving
- 19" server rack
- dSPACE Autobox
- CAN bus
- Drive-by-wire
User interface (Marble, vehicle dashboard)
- QML (QtComponents) user interface
- Declarative Marble widget
- KML ground overlay (parking spots)
- Source position (OXTS)
- Planned trajectory and target position
Range Estimation
- Electric vehicles have a limited driving range
- Energy consumption is heavily affected by
- Vehicle characteristics
- Route topology (hills, curves, speed limits)
- Driving style
- Other traffic participants
- Weather (heating, cooling, also light, ...)
- Accurate range estimation is crucial
Data Processing
- Retrieve a route [Marble]
- Analyze route
- Height profile [Marble]
- Speed limits [osmscout]
- Intersections [osmscout]
- Other traffic participants [Google live traffic]
- Simulate driving along the route
- Calculate energy consumption