British troops will today begin testing out enhanced surveillance drones, unmanned vehicles and other equipment, Defence Secretary Gavin Williamson has announced.

The Ministry of Defence say that the Autonomous Warrior experiment will last four weeks and test a range of prototype unmanned aerial and autonomous ground vehicles which aim to reduce the danger to troops during combat. The exercise will finish with a battlegroup experiment, where the best ideas and products will be tested in the toughest of simulated operational environments.

One of the key areas it is set to test, according to a news release, is the autonomous ‘last mile’ resupply. The ‘last mile’, which represents the extremely dangerous final approach to the combat zone, is crucial to ensuring soldiers have the food, fuel and ammunition to keep them alive.

Defence Secretary Gavin Williamson said:

“Our troops now have the chance to test out a huge range of robotic kit in what will be the biggest exercise of its kind in our history. We’re always working with the brightest minds in Britain and across the world to see how they can support our military of the future, but now the frontrunners have the chance to prove what they can really do on a battlefield. This equipment could revolutionise our Armed Forces, keeping them safe and giving them the edge in an increasingly unstable world.”

As well as demonstrating the vehicles during the last mile, Autonomous Warrior will also develop capabilities in surveillance which will greatly improve the effectiveness of long-range and precision targeting by service personnel.

The exercise is the result of a large collaboration between the British Army, Royal Navy, Royal Air Force, US Army, MOD, Dstl and around 50 industry participants.

0 0 vote
Article Rating
Notify of
Inline Feedbacks
View all comments
Cam Hunter

Great, but I bet these vehicles still take the same manpower to operate than what we currently use, I would get great if this capability cut out the need for large numbers of forces personel to operate and instead increase the numbers of our front line personel…

David Steeper

Cam. You understand how the Armed Forces work. I can guarantee the chiefs will find comfy seats in HQ’s for any Ruperts left without anything to do.

Daniele Mandelli

I’m just waiting it see if anything actually gets ordered.

RN and RAF have had these autonomous trials and years go by and nothing results from them until the next ” appraisal” and lump of cash to industry, comes along. And the wheel goes round and round.

On the RLC side this stuff should augment manpower not be an excuse to remove it.

The riddler

We have to have them and we have to keep laising with industry, funding it and pushing innovation otherwise they will stop and focus efforts elsewhere. Whether we purchase anything or not is irrelevant. We need them to keep pushing until something game changing comes along. This is literally DSTLs job.


Precisely. It’s pointless testing such equipment finding they are not yet ready so cancelling all development, that makes no sense at all and ignores the fact that most new innovations go through a series of hurdle crossing before becomng truly useful. Alternatively putting them into operation before they reach truly useful capabilities is costly and short sighted, freezes development and will in many cases hinder or prevent the technology ever reaching the potential they offer. That is usually less constructive than delays to useful technology.


I think the military is struggling to figure this part out. Its great if the operator is a thousand miles away performing high altitude reconnaissance etc, but the tech has become so affordable and miniaturised now, defence contractors have been marketing them at defence shows for years now where they can take off from an APC to do recon in urban environments which is great, but does it run the risk of separating the operator from their local environment where they start to rely less on their own senses? (Maybe I’m just over thinking it). I do think they will… Read more »


Not sure why people are talking about “operators” given the excersise is called “Autonomous Warrior”.
Surely this implies they are all autonomous and don’t require human operators, otherwise they’re just ‘remotely control’ devices; which have been available for decades.


Good point, in my head I’m thinking if these are targeted for the final mile resupply role, they are still going to need a pilot/driver of some description unless they plan to have a high rate of failure. Certainly if supplied by road, if by air then maybe something along the lines Amazon is developing would be possible. For recon I just don’t see how a drone would work off algorithms if it is scanning an urban area. Maybe its because I’ve watched too much Terminator and wouldn’t want that technology developed in case its next command is to automatically… Read more »


I think this is a combination of autonomous and remote control. Certainly much supply work could be carried out autonomously after an initial learning phase. Recon would no doubt have a greater remote aspect to it but I think finding the balance amongst other things is part and parcel of studies like this in real world scenarios. Think this will take years to get a trie conceptual handle on and no doubt re learnt during any early conflict phase as usual.


The US already has autonomous military helicopters that can take off and land themselves from the back of escort vessels (eg MQ-8 Fire Scout); the technology is in advance of what Amazon has.
The battlefield is going to be more challenging than an urban environment, but given we will have self-driving cars avalailable within 5 years, I’d expect it won’t be long before autonomous vehicles can handle cross-country/ off-road driving too.
It would be interesting to known if Boston Dynanics are taking part ?

David Steeper

Sean I guarantee the Army will ensure they are each commanded by an Officer with at least a dozen helpers. As for the RAF I see plenty of new berths for wing and group commanders. God bless them.

Anthony D

What I don’t get is although it takes the risk to rlc personnel out of the equation, it doesn’t stop the resupply getting shot up or disabled, which would probably result in a response and perhaps an ambush. Wouldn’t low flying or low observability drones do the supply job better?

Glass Half Full

Just to open up peoples minds about how far autonomous robots are advancing take a look at the Boston Dynamics Spot and Atlas linked below. Still lots of issues to address such as how much energy/duration can be built into some of these solutions but they are tackling some quite impressive obstacles now.

Glass Half Full

I have no idea if the comment I just wrote posted or not so here it is again. Apologies if it becomes a duplicate.

Just to open up peoples minds about how far autonomous robots are advancing take a look at the Boston Dynamics Spot and Atlas linked below. Still lots of issues to address such as how much energy/duration can be built into some of these solutions but they are tackling some quite impressive obstacles now.

www dot
www dot


Indeed I was thinking the same thing.

Separately in response to a comment above I am not clear how a drone large enough to carry useful replenishment loads is going to be less vulnerable than autonomous ground vehicals. Though in honesty it’s what testing and trials like this will no doubt attempt to fully evaluate.

Steve Taylor

1957 White Paper on Defence


The West’s reluctance to deploy armed autonomous weapons will increasingly work against us, China has no such scruples. The US navy has turned its UCAV program into a tanker program and while Russia works on autonomous T14 tanks we work on sending bottles of water to an FOB. We really need to energise our autonomous weapons program, we can leverage our world leading AI capability to protect the nation. We need an Taranis derived UCAV and submarine hunting USV and UUV’s that can be deployed by T26. Lastly the army needs something armed with anti tank weapons and a heavy… Read more »


Recent conflicts highlight the potential limitations of autonomous weapons however. We are a long way from the day they can operate effectively in the complex environments of recent wars where enemy combatant and innocent bystander are intermingled and hard to differentiate. That still needs human intellect and IMO will do for a long time to come. Plus, the ethical concerns are not groundless. Deliberately building and then permitting an autonomous machine to intentionally kill a human without human oversight is a massive thing and should not be put to one side as an inconvenient problem for ethics lecturers to worry… Read more »


Regarding the logistics side of this why spend money on something civilian industry are investing there own money into. Wait until Amazon etc have perfected delivery drones and haulage companies have autonomous trucks. Then try and get it to work. Focus on the surveillance and targeting. Real force multipliers. I am still amazed that there is no home grown Scan Eagle style UAV for the RN and Army. If I am wrong or you are aware of a current project please let me know.

john martin

No chance this will work, just look at the way it is now, we ignore resupply on all exercises as it is too difficult to do and takes lots of time and manpower. Did the job of SQMS for my tank Sqn with 4 men and a one ton rover plus a 4 ton lorry on loan from MT troop( please don’t loose it) . All this to supply 15 tanks and the rest in a war what a laugh that would have been.