Scroll to top
Get In Touch
N201C Vox Studios 1-45 Durham Street London SE11 5JH
hello@decisionLab.co.uk
Ph: +44 (0)20 3735 8580
Work Inquiries
hello@decisionLab.co.uk
Ph: +44 (0)20 3735 8580

Dstl

CREATING THE MASTER AI OF AIS THAT HELPS TO ACHIEVE OPERATIONAL SUCCESS IN THE FUTURE INTELLIGENT SHIP

CLIENT

REQUIREMENT
REQUIREMENT

DASA was seeking the latest innovations to enhance the performance of future Royal Navy ships.

CHALLENGE
CHALLENGE

Although specific AIs could be used on naval vessels, they only look after their own areas of concern. Can we make them work together for the optimal performance of a ship or fleet?

SOLUTION
SOLUTION

By using distributed deep reinforcement learning we created CIAO – the AI of AIs – to span multiple models into one powerful decision-making support capability.

Smarter choices for the Intelligent Ship
Smarter choices for the Intelligent Ship

Innovative and revolutionary technologies can help the UK maintain and enhance its military advantage. The Defence Science and Technology Laboratory (Dstl) and the Royal Navy have embarked on the Intelligent Ship programme to revolutionise its future systems by harnessing automation and artificial intelligence. This needs to work at the ship level or even fleet level to ensure that the overall optimal best course of action is taken to secure operational success.

Seeing the wood and the trees
Seeing the wood and the trees

Ships often operate in complex and dynamic environments and these days there are huge amounts of information from ships’ systems and sensors. It is hard to evaluate all of this quickly in critical situations. Individual AI applications could revolutionise many functions on naval vessels.  However, these specialist technologies will focus on only their areas of concern and optimise their particular function with no awareness of other demands in the operation of the vessel. This could lead to an overall suboptimal ship response. A way to coordinate the agents and get them to work together is needed.

Achieving ship-level optimisation
Achieving ship-level optimisation

Distributed reinforcement learning is at the forefront of AI research and holds the promise of bringing multiple AI models together to solve complex problems. We’re exploiting these concepts together with our own innovations in CIAO. It provides an autonomous decision-making capability that spans multiple functions and helps achieve ship-level optimisation.

The AI of AIs
The AI of AIs

CIAO is the AI that sits above all other AIs. A super force, it is the master of all the individual AIs that focus on their individual specific problems related to the various ship components and tasks. We use a hierarchical structure with the individual AI agents feeding into the CIAO central AI. CIAO learns how to combine the outputs of the individual AIs for any specific situation to get the best overall outcome. We’ve combined three cutting edge AI models – one for predicting engine failure, one for defending against incoming missiles and another to avoid and deter small vessels swarming and attacking it – within CIAO to get the best of each AI.

A tool that resolves conflicting choices
A tool that resolves conflicting choices

We have developed CIAO so that is flexible and extendable. It can be adapted and applied to similar problems requiring reactive decision models where conflicting choices need to be resolved in real time. CIAO lays the foundation for a capability that can address a wide range of needs within the Intelligent Ship concept, and it could have wider application across defence and beyond.

We’re a team of innovators who are excited about unique ideas and help companies to create amazing solutions.

Dstl

We use cookies to give you the best experience.