One of the biggest attractions is cost – the software can use any virtual reality (VR) headset which can cost upwards of a few hundred pounds. Gathering hundreds of soldiers to blow up a bridge is very expensive, particularly compared to donning a headset.
According to research by EY, the training could save the MoD some £1.3bn.
“Training in real environments is expensive. This is a way to train across a much broader range of environments, and much more quickly,” says Prof Jordan Giddings at University College London, who is also a particle physicist working on Improbable Defence’s project.
Robinson hopes it will also help speed up training to the minute. “The old adage is that the military is very good at fighting the last war, as opposed to the current conflict. And what our technology enables is, is a very rapid update of that data and that information.”
Information from the current war in Ukraine can be plugged into the system, for instance, using videos captured by civilians on phones, as well as from media reports and satellite images.
Robinson says this could help everything from understanding new tactics to how to apply anti-tank weapons and use drones. “The amount that we’re learning from a tactics and techniques and procedures perspective, from watching, you know, this awful conflict unfold in front of our eyes is extraordinary.”
Much of the work involves gathering together the kind of data the military may not have realised it could use, he adds: “We are working with some of the top academics in the UK, who have a specific expertise, like, for example, population movement in urban areas. And that’s something that traditionally you wouldn’t get out of the defence industry.”
Improbable is one of many companies developing VR training. Farnborough-headquartered Qinetiq, better known for its robotics research, is also developing a successor to its training programme which trained 16,000 British troops in preparation for deployment in Iraq and Afghanistan.
BAE Systems has also created a simulation for landing its F-35 jet on an aircraft carrier, mimicking the volatile pockets of air which pilots will have to contend with if they are to make a successful landing.
But there is also much potential to be tapped, adds Prof Jordan. As with video games, such environments can create a whole world with all of its complexity and randomness.
Improbable’s Robinson argues that the software’s potential lies in its ability to become part of a broader simulacrum that everyone from politicians to industry leaders can use to try and solve some of the world’s problems such as climate and energy.
Using artificial intelligence, much like the new breed of autonomous car that is in development, the software can also learn and start to offer decisions itself.
This intelligence can then be integrated into some of the computerised control systems used by drones, jets and modern armoured vehicles to make some of the smaller decisions in a battle.
It can also be used for rehearsals in the field, just before a mission, and can ultimately be married up with the real world as an augmented reality, where soldiers can be carrying a gun with blank ammunition around a field with simulated civilians and challenges, blending virtual and field training.
The technology “really isn’t just for training anymore” says David Taylor at Qinetiq. It can also be used to visualise data and test designs of new military hardware.
For now, a giant integrated system mimicking the whole world is some way off. A nearer challenge is making the simulation as close as possible to real life, to the point where empathy and compassion can be felt in order to make good decisions.
“They’re all components in that environment, you’ve got to make it look and feel as real as possible,” says Robinson.
