“These systems often require great physical effort to run and only consider a limited set of scenarios, resulting in significant uncertainty regarding outcomes.”
The Defence Science and Technology Laboratory (Dstl) is offering £400,000 to UK enterprises that can come up with new innovation in machine learning that can help model future conflicts.
UK-based Digital Catapult work on emerging technologies for the public and private sectors. In 2017 they engaged with 638 start-ups and worked within 42 industrial collaborations.
Digital Catapult has put out an open call and is organising a pit stop event, similar to a hackathon, to seek innovative ideas for the use of artificial intelligence (AI) and machine learning in future Dstl decision-support combat models.
As part of their role in the open call Digital Catapult will review all applications, vet and invite potential attendees.
Dstl principal analyst Mark Gould commented in the announcement that: “By working with Digital Catapult, Dstl will expand its supplier base to include up and coming small and medium enterprises who are developing exciting innovations in AI and modelling but who don’t currently work with defence.
“These new suppliers offer us opportunities not only to expand our supplier base but also to explore how innovations being used outside of defence can be used to improve the support we offer to MOD into the future.”
Model Future Conflicts
During the conclusion of the Pit Stop event, participants will be invited to submit proposals to Dstl for follow-on studies, investigations and/or developments arising from the discussions and activities in the event.
The highest amount that a contract can be awarded is £400,000.
Currently, most of the combat modelling techniques of the Ministry of Defence either represent command decisions through rules-based systems or rely on human users to provide the command context, while simulation handles the complex conclusions of the outcome.
These systems require physical effort to run and only consider a limited set of scenarios, resulting in uncertainty regarding outcomes.
As part of ongoing work to develop a suite of capabilities that can reduce the manpower needed to make credible command decisions and enable the simulations to consider a wider range of situations, Dstl is seeking to understand how AI can be used in this mission.
The closing date for the applications and registration is the 15th of October 2018.