Dstl Pit Stop: Intelligent Ship

How can artificial intelligence (AI) and automation assist the Royal Navy in achieving mission success?

At this Pit Stop, startups and innovators will have the unique opportunity to collaborate with and inform Defence Science and Technology Laboratory (Dstl) as we look to demonstrate how the application of advanced digital technologies can transform the Defence sector. This project sets out to revolutionise future ship design with the application of AI and automation in the digital space, optimising data usage to make better informed decisions.

The challenge

The Dstl project, Intelligent Ship, is working towards revolutionising ship design and development for the Royal Navy (RN). This Pit Stop seeks to collaborate with Dstl and a collection of innovators to help drive forward the application of advanced digital technologies for Intelligent Ship.

The project aims to address how best to make use of automation and artificial intelligence (AI) applications to ensure optimal use of all the available data. This includes the examination of how operations could be better undertaken through the use of automated techniques for data processing, planning and decision making.

Current RN ships have a variety of diverse and isolated data streams; this mix of information and the need to integrate and make timely decisions based upon it creates a highly challenging and potentially stressful environmental for RN command teams. Data has to be prioritised and understood whilst providing situational awareness, allowing the command team to make life-saving decisions.

The interaction between humans and machines will be a crucial to success, with an aim to build on the strengths that each brings to the mission. To ensure decision confidence, applied AI solutions should be transparent and clear; so that the humans understand and trust the outputs, and can conduct meaningful post mission analysis.

This task is limited to the systems within a single naval platform. However, Intelligent Ship will be required to work within a defence context to form part of a task group, squad or swarm of platforms co-operating to achieve successful mission outcomes. Therefore, the Intelligent Ship is being designed to operate independently or as a team.

Below is an example scenario that the future Intelligent Ship system may be involved in.

The Scenario

The Royal Navy’s latest innovative ‘’smartship’’ HMS Perceptive is deployed, conducting operations away from UK waters. It has an Intelligent Core, comprising of AI supported decision aids and advanced automation. The aim is to assist the Command team in making their decisions, utilising numerous information sources including (but not limited to) on-board, off-board and networked sensors.

HMS Perceptive is on a patrol mission taking her through a channel (natural choke point) between two land masses. The Command team require the most complete understanding of the recognised picture around them, across all environments and spectrums to make effective decisions.

HMS Perceptive is tracking all detected surface and air contacts within sensor range, and monitoring the underwater space. Amongst all the traffic, she has identified 10 individual, separate, challenging air, surface and sub-surface contacts. They are all dispersed, have separate origins, yet have courses which converge (or come close) with HMS Perceptive. This behaviour seems to be suspicious, not seen in regular Patterns of Life (PoL) and is potentially threatening.

The challenge for the Command team is to ascertain the intent of these contacts prioritise responses to decide if they are exhibiting normal behaviour or pose a potential threat. This decision making process should be aided by the Intelligent Core.

Who should apply

We’re seeing early-stage companies, startups experts, and research groups that are in the process of developing new cutting-edge technologies.

Areas of advanced digital technology applications and solutions s include (but are not limited to):

• Autonomous vehicles
• Artificial intelligence/machine learning
• Human-machine teaming/virtual assistants
• Data visualisation
• Data integration, processing and analytics
• Machine vision
• Automation
• Big data and analytics
• Simulation and modelling
• Digital Twins
• Mission planning
• Decision making and supporting decision makers
• Adversarial learning
• Uncertainty in ML predictions
• Explainable AI: understanding how the system reaches its conclusion

We may also consider energy management, cyber security and future networks (e.g. 5G, LPWAN) as a part of the above areas.

We are NOT interested in:

• Autonomous lethal weapons systems[1]
• Novel approaches that don’t meet the brief