"Current computer software, including operating systems, have poor reliability and security. Unfortunately, this lack of reliability also applies to embedded systems, that is physical systems controlled by a microprocessor, such as those used on cars, airplanes, MRI scanners, pacemakers, insulin pumps, cameras, set-top boxes, mobile telephones, appliances, and many more. Software failures in these devices can have catastrophic effects. Bugs in the code of the Therac-25 and Cobalt-60 radiation machines killed several dozen people due to overdoses of radiation and bugs in the software of the Prius’ software-controlled braking system have led to fatal accidents, for example.The related ERC Advanced Grant was about redesigning operating systems to make them more reliable, with a focus on desktop systems.
In this project we want to use the results of the research to try to make embedded systems more reliable and then move toward commercializing the software. One area of interest to us is embedded systems that are currently or will soon be networked. It will not be too long before, say, next generation pacemakers get networking capability to send patient monitoring information to the hospital in real time. Such systems need to be extremely reliable and protected against hackers. They also need to be updatable while they are running. No current embedded systems provide this kind of capability, but we think we can provide it and we believe there will be a market for it. The first step is to port our desktop system to the ARM processor, which dominates the embedded systems world, and make a prototype to show potentially interested parties. At that point we will be able to see if the project is commercial viable. Before the ARM port and a sample application have been done, that is nearly impossible to do."
Fields of science
Call for proposal
See other projects for this call