Self-Healing Computers For NASA Spacecraft

Roland Piquepaille writes "As you can guess, hardwired computer systems are much faster than general-purpose ones because they are designed to do a single task. But when they fail, they need to be totally reconfigured. This can be just a costly problem in a lab on Earth, but it can be vital in space. This is why a University of Arizona (UA) team is working with NASA to design self-healing computer systems for spacecraft. The UA engineers are working on hybrid hardware/software systems using Field Programmable Gate Arrays (FPGAs) to develop these reconfigurable processing systems. As the lead researcher said, 'Our objective is to go beyond predicting a fault to using a self-healing system to fix the predicted fault before it occurs.'"

Re:The 9000 Series has a perfect operational recor

[Alpha Whisky]

Actually we do have very effective lip reading computers http://en.wikipedia.org/wiki/Automated_Lip_Reading [wikipedia.org] they just don't understand what they are reading. The documentary about lip reading the silent movies of Hitler was very interesting from a technical standpoint, even if it did turn out that they had hours of recordings of Nazis making small talk about the weather.

Doesn't this already exist?

[flnca]

What will Starbridge Systems [starbridgesystems.com] think about that? Didn't they develop a dynamically reconfigurable computer that ran Windows NT as a test application on 10,000+ FPGAs back in the 90ies? IIRC, they also had a software framework able to automatically implement software fragments in hardware using FPGA auto-configuration.

Self-repairing computer systems for spacecraft have been in the discussion for decades, and every now and then we get hear about a new project. This project certainly is a good idea, hopefully it will work.

BTW, Motorola (now Freescale) developed self-repairing processors for military applications a couple of years ago.

Re:Reconfigurable Computing, Fault Tolerance

[arktemplar]

I had mentioned this some time back as well, but polymorphic processors like MOLEN(tu delft is doing this one), might be usefull for this sort of stuff. The theory behind it is simple, and extends to modern multicore systems as well basically break up the instruction set into microinstructions (all processors that I know do this part), then have any one of the many computational units available do whatever work is required in order to implement those microinstructions. the translation is done by the core processor it self, it can be made redundant etc. as required, they already have an FPGA implementation of it and are using it for research into super computers.

Re:Not new

[RiotingPacifist]

I wonder if it will every be cost effective to put FPGAs in consumer systems, i can see them really helping in bottlenecks (why waste cpu on doing the same processing over and over, ship it of to a specialised FPGA) and in low power situations (why wake up the cpu when you can program the FPGA to do 50% of the wake-ups), unfortnatly i can only see this helping mac & linux, as the windows kernel being closed makes implementing this stuff down to MS not the chip makers.

I think the goal of this project isnt high performance, but high redundancy, there are only so many backups you can put on a probe, with this if they do it right you could end up with a system where any core could break and be fix/replaced at fraction of the cost shipping 2 chips for every component. Unfortunately often the sensors go before the core technology so i dont know how effective that will be, plus due to a lack of funding often projects are abandoned long before all the systems are broken