DASA/Dstl

Hadean has successfully delivered a new training capability funded through the Defence and Security Accelerator (DASA) under a Defence Science and Technology Laboratory (Dstl) contract.This project enabled us to research and develop a scalable, AI-driven synthetic training environment designed to populate the digital worlds of military training with realistic and dynamic human terrains.

Download Case Study

Scaling a representative AI-powered synthetic training environment

The project uses AI to create a realistic simulation of A3E (Audiences, Actors, Adversaries, and Enemies). The system features end-to-end AI interfacing, responding dynamically to events and actions in the training
exercise, with end-to-end AI interfacing, as well as in-depth after-action analysis.

Our solution injects civilian behaviours and sentiment into a dynamic social
media environment, automatically summarising events for after-action review (AAR). This integration capability provides a more authentic and complex training experience for trainees, enhancing their ability to navigate multi-domain operations.

Responsive synthetic human terrain for Live, Virtual & Constructive (LVC) training

Real-Time Civilian Behaviours and Sentiment Tracking in Dynamic Environments

Automatically summarised events for after-action review

Responsive synthetic human terrain for Live, Virtual & Constructive (LVC) training

Real-Time Civilian Behaviours and Sentiment Tracking in Dynamic Environments

Automatically summarised events for after-action review

Hadean's Role

Hadean’s AI-powered spatial computing provides a scalable Pattern of Life simulation for dynamic human terrains, integrating with large-scale SSEs and LVC training for enhanced multi-domain exercises, situational awareness, and decision support.
01.

Dynamic A3E Entities:

Our simulation created responsive A3E entities that adapt to cues, stressors, and frictions across human and information domains.

02.

Scalable Simulated Populations

We implemented the ability to scale populations of simulated civilians according to the specific requirements of each simulation scenario.

03.

Real-Time Social Media Sentiment Tracking

A large language model (LLM)-generated social media feed reflected population sentiment, tracking civilian responses to events in real-time.

04.

Immediate After-Action AI Analysis

Our AI provided immediate insights through after-action reviews, enabling faster and more informed decision-making.

05.

AI Analysis:

Our AI provided immediate insights through after-action reviews, enabling faster and more informed decision-making.

Hadean's TRL6 demonstration of an LLM-powered training capability for the DASA competition, Populating the World of Training, was an excellent showcase of UK innovation applied to defence - Hadean delivered as good an outcome as I've seen.
Giles Moore

Principal Scientist, Futures and Innovation Group, Dstl