Why not analyze measurement data from automobiles, mobility infrastructures and mobile phones and use it for autonomous driving? Precisely this idea is being implemented by SADA (smart adaptive data aggregation), a joint project involving fortiss. The results of the project flowed into the development of “EO smart connecting car 2”, a robotics-based electric car. The demonstrator can be viewed during the Hannover Trade Fair (April 23-27).
Today, modern automobiles and mobility infrastructures are both equipped with numerous sensors, not to mention that the drivers’ smart phones are in essence nothing more than portable measurement sensors. If this heterogeneous data were combined and analyzed in real-time, this so-called data fusion could offer extensive benefits, including the creation of traffic prognoses, optimized traffic flows, better road utilization and increased safety. It would thus help avoid traffic congestion, sink energy consumption and reduce CO2 emissions.
The issue is that the data from mobile devices and mobility infrastructures does not lend itself to modular use, nor can it be coupled with the required flexibility, since the sensors and analysis functions, as well as the hardware and software, stem from various manufacturers, which impedes communications between the components.
Autonomous data fusion
The participants in the SADA project have now developed a solution: a system that analyzes, merges and intelligently links sensor data in real-time from stationary road infrastructure components, automobiles and smart phones. The basis of the system is a high-speed communications architecture that connects the heterogeneous sensors.
The system recognizes which data is available and selects the information relevant to the respective traffic situation. Researchers at fortiss developed the corresponding data fusion algorithms, in addition to a run-time system for the data exchange. A centralized, back-end system processes and administers the data.
Improved driver assistance systems
The SADA system can do more than improve driver assistance systems. It can also be used to support autonomous vehicles. Because the sensors also capture information about the surrounding environment, the data can be used to generate acoustic warnings when pedestrians cross the street or avoid unnecessary braking maneuvers by anticipating traffic situations or road conditions. Electric vehicles can be equipped with new features such as operating range estimates, energy-saving driving recommendations or autonomous driving to the next charging station.
Apart from fortiss, other participants in the SADA project include DFKI Robotics Innovation Center, BASELABS GmbH, NXP Semiconductors and ALL4IP TECHNOLOGIES GmbH & Co. KG. Siemens AG is leading the consortium. The project was funded though a three-year (2015-2018), €3.8 million grant from the German Federal Ministry for Economic Affairs and Energy.