In the ongoing conflict in Ukraine, we see a significant use of aerial drones for reconnaissance and combat roles, but there seems to be a noticeable absence of ground-based robots, especially those that could potentially carry out direct combat tasks, such as armed remote-controlled robots. Given the advancements in robotics and autonomous systems, this raises several questions:

1.Is it a matter of terrain navigability, operational reliability, or something else like logistic?

2.Are drones simply more cost-effective and versatile for the needs of modern warfare?

3.Is it a matter of ethics and reduced escalation?

4.Despite these challenges, are there instances or documented cases where ground-based robots are being used in the Ukraine conflict or elsewhere for combat roles? If so, what types and to what extent?

  • 9
    $\begingroup$ Have ground-based robots ever seen widespread deployment in any conflict? I can't think of one. $\endgroup$
    – F1Krazy
    Commented Feb 26 at 14:19
  • 4
    $\begingroup$ My guess: navigating in the air is much more simple. Why walking if you can fly. $\endgroup$ Commented Feb 26 at 15:12
  • 1
    $\begingroup$ There's Engineering SX. $\endgroup$ Commented Feb 26 at 15:16
  • 1
    $\begingroup$ Ground tactical robots are hard to do: wheels don't work well in rubble or in buildings, so research has been going into biped/quadrupeds or even snake bodies. Not at all trivial compared to flying or even boating. In addition armed air drones are a natural extension from surveillance ones. Ground robots in armed forces have been researched/used for logistics (self driving trucks were the goal of the initial 2004 Darpa challenge) and bomb disposal. Lethaler stuff just ain't ready, right now, for Ukraine, not yet. $\endgroup$
    – Italian Philosophers 4 Monica
    Commented Feb 27 at 6:06
  • 1
    $\begingroup$ This appears to be a question about technical practicability, not about the political implications of ground-based drones. So this question doesn't belong here. Berhaps Engineering SE can help answering this question. I will migrate it. $\endgroup$
    – Philipp
    Commented Feb 29 at 12:41

4 Answers 4


Insufficient technology. Ground drones have much more difficult mobility, navigation, and control challenges due to varied terrain and obstacles which not only impede mobility but radio communications. Just operating like that in an area without threats would already require an unprecedented level of automation and mobility.

But these drones are not operating in areas without threats, and unlike airborne units, they are much less survivable since they present targets that are not only easy and cheap to destroy, but to capture intact.

That means the drones need situational awareness to detect and analyze threats, further enhanced automation to know how to evade or hide, and further enhanced mobility to actually do so. I think the Mars Rovers might be the only robots in actual use that come closest to satisfying these criteria...and they obviously aren't going to cut it.

The most passive solution to this is to reduce size (like mouse to housecat sized) which eases intelligence aspects, but that in itself is a serious technological challenge and limits capability. The other approach is superior environmental awareness and judgement response in a larger package but that's how you get Terminators.

At the moment, it seems difficult to beat a small remote control truck with explosives tied to it. Lacks all of the above except for small size but makes up for it by being dirt cheap and worthless to capture.


Ground-based robots are still a nascent thing. Boston Dynamics (robot company) has been building ground-based military robots for a while and they still have problems with them

Kyle Olson, a spokesman for the US marine corps’ Warfighting Lab, explained the problem to Military.com: “As marines were using it, there was the challenge of seeing the potential possibility because of the limitations of the robot itself. They took it as it was: a loud robot that’s going to give away their position.”

The problem there was noise. You can build a much quieter battery-only model... but that impacts its ability to carry loads and range. From what I have seen of the machine itself, terrain wasn't an issue. And these machines are very complex.

Israel has some ground drones they actively use, but they are very expensive

A vehicle alone costs approximately $600,000. With the operating system, the price runs to several million dollars, depending on what equipment is installed on the robot.

Ground robots have to contend with both terrain and enemy action (taking fire, land mines, etc.). They are also no faster than any land vehicle. Given their cost, nobody is using them on a wide scale.

Drones are simpler, and thus easier

What is making drones a far more effective 21st century warfare tool is

  1. They're much quieter
  2. They're much smaller (much harder to spot on radar)
  3. They're cheaper to make
  4. They can be expendable
  5. There's not as much in the way of defending against them

Mind you, I'm not talking about a $30M MQ9 Reaper, I'm talking much simpler drones. I mean, you can now build flying bombs that can reach hundreds of miles

Drones have reached the Moscow region, which is about 450km (280 miles) from the border with Ukraine.

This lets you attack targets on the cheap. Ukraine definitely hit on something there

The strike set at least two military planes on fire, and a satellite image taken of Pskov Airport after the incident shows two aircraft completely destroyed.

  • 2
    $\begingroup$ The bigger drones from the last quote are basically just cruise missiles, albeit built cheaply/COTS. Those have otherwise existed for decades. What made the smaller drones practical is huge advances in battery tech, camera tech, and some control circuitry miniaturization, including gyro sensors--flying smartphones basically. $\endgroup$ Commented Feb 26 at 16:05
  • $\begingroup$ Lo and behold when this gets miniaturized enough. We're gonna see "guided bullets" that one can fire up in the air and automatically tracks and kills someone miles away. "Ai" basically. $\endgroup$ Commented Feb 26 at 17:59

Well, there were some ground 'drones' (like used to sneak-up mines) but...

Let me add a few more elements why I think this is not widespread.

Small air-based drones (especially suicide-FPV and grenade droppers) have become common in that war because they 'fly under the radar' of the bigger SAM systems. And even if they'd get detected it's economically disadvantageous to shoot them down with a big/expensive missile.

A small ground-based drone 'sees' huge ground obstacles relative to its own size. Every pothole is giant crater and every boulder is a mountain for it. So if you make a ground-based drone small it has more terrain navigations difficulties for that reason alone.

Speed is also a factor here. Air-based drones can approach and deliver munitions (or blow themselves up) before they can be shot down. Sometimes soldiers do shoot at air drones with their individual weapons, but the small-air based ones are fairly hard to hit because of distance and their speed. Speed also has another advantage: there are for instance videos of FPV drone catching up with vehicles running on roads etc. at 'full speed'. Ground based small drones are slow, and mostly rely on not being seen/detected before it's too late (as in that video).

As for the 'medium-sized' ones... Both sides have experimented with these too. Russia even has an official programme for unmanned [ground] combat vehicles. Ukraine has experimented at least with evacuation/carrier ones. But the larger you make them, the easier they are to see and shoot like any other vehicles, so there's that tradeoff. And Russia's own evaluation of that program was that's not quite ready for prime time. N.B. It looks like Russia actually deployed some of their medium-sized ground drones in Ukraine, but they didn't fare that well against [flying] FPVs.

  • $\begingroup$ I've been dealing with house mice and trying to get at them deeper and deeper into walls and crawlspaces and have concluded a big reason they can climb anything is that they weigh almost nothing so require almost no force to hang onto things and have more footholds that can support them. Probably even more important than a UAV because in UAVs you can often cheat by making the wings bigger, fly faster, or rotor spin faster. Cats are like giant mice that can't climb. Dog mobility would be the bare minimum. $\endgroup$
    – DKNguyen
    Commented Mar 1 at 16:25
  • $\begingroup$ @DKNguyen: Somewhat related: US army used/evaluated 'robot' dogs for tunnel warfare. I guess Israel too. $\endgroup$ Commented Mar 1 at 17:25
  • $\begingroup$ Because cats have superior mobility dogs...but they're cats and won't listen to you lol. I believe there is also that old story of USSR anti-tank dogs. $\endgroup$
    – DKNguyen
    Commented Mar 1 at 19:29

Robots have become common for demining.

Here is a design that was in service in the AFRF before the start of the war - the Uran-6 (image: topwar.ru) uran-6

And here is its mid-2023 counterpart, the MGR-4. Adoption status unclear. (image: topwar.ru) mgr-4

The evolution here was not in function, but rather production cost. The first one was a peacetime military design. The second is built in wartime, on the chassis of a civilian construction vehicle (something roughly like a Bobcat), for a fraction of the cost.

A demining robot is of course slow and impossible to hide due to dirt flying everywhere and noise of explosions around it. It would not last in a frontline application. But it's another an example of the trend for cheap-and-cheerful rather than high performance.

  • $\begingroup$ Yeah, but AFAICT not in a combat/lead situation. Maybe you want to add that. At least Russia still generally leads with a tank with a plough. And IIRC so did Ukraine, although they've not attacked much in recent months so my mem is less fresh on those vids. $\endgroup$ Commented Mar 5 at 6:35
  • $\begingroup$ @Dolphin - Yep. These are for anti-personnel mines. Relatively small 4-6 tons. $\endgroup$
    – Pete W
    Commented Mar 5 at 12:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.