Want to help the MOD separate submarine signal from noise?
The challenge: stripping "defective" data streams prior to processing
As public sector tenders go, the Ministry of Defence's latest £530,000 opportunity isn't the world's biggest. But for an ambitious data wrangler, it may prove an intriguing one. MOD is after someone to help it separate signal from sensor noise (time-series data from large arrays of submarine sensors capturing noise and vibration data.)
The job will involve developing and demonstrating "algorithms that can be used for identifying and classifying defective channels so that they can be removed from the data stream prior to processing by other algorithms, as part of a larger data analysis system" MOD says in a tender for the 8-month contract, posted October 12.
Exclusive: We reveal NATO’s first ever CIO hire
"The algorithms must be suitable for insertion into an open system and where they depend on the use of training data, a strategy for obtaining that data will need to be devised and implemented.
"A validation and verification strategy will need to be devised and the success of the algorithm(s) in identifying defective channels demonstrated" it adds. Submissions/request to participate need to be in by November 15. Potential suppliers will need to be compliant with the Government’s Cyber Essentials Scheme.
The security classification is OFFICIAL-SENSITIVE. Details on how to apply are here.
The tender comes after a US nuclear submarine struck a mystery object its sensors failed to pick up, in the South China Sea on October 2. In other submarine technology news, Rubin Design Bureau, a Russian-based defense contractor that designs nuclear submarines for the Russian Federation’s Navy, was hit by a cyber attack this year after a spear phishing campaign, tracked by Cyberreason, which attributed it to a Chinese APT.
