Future applications for automobile systems will require much higher levels of configurational flexibility and scalability. Even though about 80% of all innovations in the automotive industry are related to their embedded electronic systems, today's mostly static systems cannot handle new technical challenges.
That's because the configuration is defined during the development process, and it remains stable over the vehicle's lifetime. Compounding the issue, the typical embedded components in vehicles consist of decentralized electronic control units (ECUs) based on proprietary hardware and software components. Usually, they're inseparable functional units.
In the future, vehicles will be able to modify their behavior to suit their environment. One option includes mobile communication and entertainment devices. Phones, PDAs, and media players will be integrated and networked in an ad-hoc fashion within the automobile. By offering location-aware capabilities, the vehicle will be able to recognize its surroundings and then enable communications services to come and go depending on network connection availability.
Internet bandwidth would increase as soon as the car enters a wireless local-area network (WLAN) hotspot, automatically activating services capable of exploiting the bandwidth (such as downloading software updates). Once the car leaves the hotspot, the attached mobile phone's bandwidth would return to normal, and all services requiring the high-speed connection would be disabled. In addition, updating content such as music and movies and synchronizing a user's calendar could occur overnight when the vehicle is parked at home.
To provide these and many other capabilities, a new way of developing automotive systems is evolving. The Autonomic Computing Paradigm is a shift toward software-centric and network-centric ECUs that will support these new dynamic systems. No longer will software be an ECU adjunct. Instead, it is becoming the dominant component of electronic automotive features. The ?under the hood software? will be upgradable to some extent. Also, functions will be able to be moved from one CPU to another to prevent possible hardware failures. This introduces a cost-efficient dual redundancy the industry hasn't afforded before.
The DYSCAS consortium is spearheading research to develop these future dynamic automobile electronic systems. Members include BOSCH, Enea, various universities, and OEMs like DaimlerChrysler and Volvo.
Standards like Autosar (Automotive Open System Architecture) also will play an important role as standardized hardware/software component architectures and infrastructures replace proprietary solutions. Open middleware level interfaces will let third parties develop functions that can be deployed across many platforms and configurations, regardless of automotive manufacturer, brand, and model.
The Autonomic Computing Paradigm seems to be the right approach to solve the upcoming crisis in auto technology complexity. It advocates a self-managing behavior in which applications modify their behavior to suit their environment and context. The autonomic approach reduces the emphasis on the pre-configuration of components. Instead, it relies on inbuilt learning and discovery capabilities combining software and the topology of networked sensors.
