Nowadays microprocessors are used in thousands of items that were previously not computer-related. These are embedded inside such devices. Together with the proprietary software controlling them, they each form a so-called embedded system. Several of these are at work in an average middle-class household, hundreds in cars, and as the American multinational semiconductor company AMD concludes in its 2014 annual report: “There is significant demand […] which address the growth of data and content in a world of 50 billion connected devices”.
But embedded software is not keeping up with the speed of this development in terms of safety and quality. “We do not understand what the software does, regardless of how well educated or smart we are”, says Holger Hermanns, Professor of Dependable Systems and Software at Saarland University. He points out that proprietary software has become an opaque layer between functionality and users. “This kind of embedded software locks us out of the products we own”, Hermanns says.
Hermanns wants to change this by investigating the embedded software which operates the batteries of mobile devices such as smartwatches, smartphones, or robots. As part of his new project, awarded by the European Research Council, he wants to make traditional software development processes for proprietary power management software more transparent, develop new software tools, and make these open source. Hermanns explains “What makes dealing with the battery so interesting and necessary are the following two conflicting goals from the user perspective: First, the user wants to be in control of his device; secondly, the device should not endanger anyone.” The latter can occur if the battery is used incorrectly. This can lead to injuries caused by overheating, leaking, or the device catching fire.
As a domain of choice Hermanns selected e-bikes, whose transformation in the field of embedded systems has just started. “This is a good opportunity to do things differently because the consequences are now becoming more apparent… a huge share of bikes are maintained by their owners and there is a certain tradition of modularity and interoperability of components which are often inexpensive”, Hermann says. Already-existing devices like electronically controlled gear shifts and an electronic suspension confirm that observation. “Both are safety critical”, Hermann states, “and especially fiddling with e-suspension software does not seem like a good idea, unless the tinkerer is supported by tools helping him to understand the physical implications and software limitations.”
Hence, Hermanns and his research colleagues are working closely with the committee developing the standard “EnergyBus”. This is an open standard to ensure the smooth interoperation between all electrical components on e-bikes or comparable electric vehicles, such as batteries, chargers, sensors, motors, or control interfaces. The EnergyBus forms the foundation for a new world standard prepared by ISO / IEC. According to that standard, manufacturers have to guarantee that their battery drives each electric bike through a standardized plug and standardized software.
In order to make component producers comply with the quality and safety requirements of this standard and the related software, Hermanns uses “quantitative verification”. This is a branch of computer science that explores fundamental questions for embedded systems and develops software tools which are also tested in real life. In this way one can review automatically, and therefore guarantee, important aspects of embedded systems functionality (such as the air bag opening only in case of a collision). For software controlling the power supply to an electric bike, these are, for example, operational safety (accidental overloading does not lead to long-term damage), reliability (with normal use, the battery lasts for more than four and a half years), and performance (in 80 percent of all charging operations, 20 minutes suffice to drive 20 kilometers).
Hermanns is convinced that ordinary battery technology should support the owners of such devices and not limit them, regardless of whether a smartphone or an e-bike is being used. Hence, he named his project “Power to the People. Verified.” According to Hermanns, this stands for two objectives: first, to assure a power supply for all mobile, embedded systems, and secondly, not to disenfranchise citizens using technical equipment any longer. The European Research Council is granting his project 2.4 million Euros over five years.
Computer Science on the Saarland University Campus
Apart from the Saarland University Department of Computer Science and the Cluster of Excellence on “Multimodal Computing and Interaction”, there are other research institutes which have a global reputation and are based at the Saarland University campus. These are the German Research Center for Artificial Intelligence, the Max Planck Institute for Computer Science, the Max Planck Institute for Software Systems, the Center for Bioinformatics, the Intel Visual Computing Institute, and the Center for IT-Security, Privacy and Accountability (CISPA).
For further information please contact:
Professor Dr. Holger Hermanns
Chair for Dependable Systems and Software
Phone: +49 681 302-5631