Marines’ LS3 robotic mule is too loud for real-world combat | Hacker News Reader

Marines’ LS3 robotic mule is too loud for real-world combat

arstechnica.com

37 pointsrfreytag39 comments

A friend of mine who was in the Marines ROTC said he had seen one of these being demonstrated a few years ago and I got the chance to ask him about it briefly. He was scant on details but did say there was no chance they would ever use it because it was too loud. I replied that maybe they could make the engine quieter and he said it would still be too loud. He also said that there were limited useful applications and they just wouldn't really need robotic pack mules that much. I was a doubter but I guess he saw this coming from a mile away.

I never heard from the military if the heat signature was a problem as well. The article sights noise, but that thing has to stand out on IR goggles or similar monitoring devices.

Good point, I would imagine that would be concerning but most of our current enemy combatants probably cannot afford heat sensitive googles. I wonder what the radar signature is and if it can be distinguished from a cow?

It would be interesting if they could be used as scouts.

most of our current enemy combatants probably cannot afford heat sensitive googles

This is likely to change as electronics get ever cheaper.

So, the conclusion is that there is a market for a really quiet gas powered engine. I was thinking when the first prototype came out that working on a quiet engine was what they were going to spend the money on, the basic robot seemed to be ok for what they needed.

I'm curious why that problem would not yield to a concerted effort. I spent some time quieting my generator motor which consisted primarily of catching and slowing the exhaust gases I stopped refining that solution when I could buy a cheap (and reasonably quiet) generator from Yamaha. For three or four million I'd be happy to restart my efforts on this project :-)

Not sure what the required power output is for a robot of this type, but I was impressed by this tiny 400W gas turbine: https://www.youtube.com/watch?v=Al8elCF816g

That is a pretty cool motor, but the basic physics of a robot of this size suggests closer to 30HP or 22,500W. The fuel cell decomposing idea is an interesting one too. The good news is that the customer is price insensitive so you can charge a lot to cover your R&D expenses.

Seems like an possible application of hydrogen fuel cell tech, sufficiently miniaturized. IINM, there are systems that can use gasoline or ethanol as the source of H+.

Boston Dynamics, the prime contractor for LS3, has bowed out of the military robotics business since being acquired by Google two years ago.

One feels that this might also have been a major factor, although perhaps it was easier to pin the decision on a technical issue.

Or there's always this 1.0 variant. Has a known tendency to kick and bite. But usually quieter. Can haul almost as much.

https://en.wikipedia.org/wiki/Mule

For just $12.99 you can get the CD: https://en.wikipedia.org/wiki/Gov%27t_Mule

Or M.U.L.E., made of scrap parts: https://en.wikipedia.org/wiki/M.U.L.E.

Or this one. The alternative fuel source makes it slightly quieter:

https://www.youtube.com/watch?v=VXJZVZFRFJc#t=23s

Considering the part of world we're mostly fighting in now, maybe it's high time to reactivate this Civil War era experimental unit:

https://en.wikipedia.org/wiki/United_States_Camel_Corps

Unfortunately Mule 1.0 suffers from reduced payload at higher altitudes ( falling below 100kg for the top breed ) and tire quickly, requiring frequent resets.

They also require 30 to 50 litres of water per day, depending upon climate, so on a three-day mountain patrol they can basically carry only their own water and no other payload. Not terribly useful. Plus fodder...

Courtesy of the Argentinian army, here's a table of the mule resources required to sustain a 171-man company on a six-day mountain march:

http://fmso.leavenworth.army.mil/documents/groundcombat/char...

Yes, 232 mules and 171 men.

Marines don't even speak while on patrol, it's pretty obvious that a robot powered by a gas engine wouldn't be a good idea.

MARSOC units apparently use electric dirtbikes to quickly through rough terrain on some missions.

...which suggests a possible solution to the current dilemma.

I don't know...I can't imagine a small electric dirt bike being useful with 100 - 120 lbs of gear (typical infantryman payload with full pack), plus the rider. I think the technology needs to come a long way.

I could have been clearer. I meant they could pull the electric motor out of the dirt bike and put it into the robot.

I doubt that the mule would be of much use then. The range of the fully loaded mule was fairly limited. A battery powered version would have an even more limited range. This is possibly true even with a fuel cell.

I don't think that it's a total waste, provided the Government gets the R&D. The machine can follow a human without direct controll. This could be added to alternative tech.

At the same time it could be useful in non war scenarios like fire fighting or rescue and recovery. In both situations stealth is not important. Getting into remote areas might be possible, and better with these machines. If necessary, depending on environmental factors, fuel deposit could be dropped ahead of time.

Something that can transport a grown human quickly, can probably also transport a roughly equivalent weight more slowly? Or at least it will be able to do so given enough R&D? Electric motorcycles have been available for several years, from a variety of companies. Either the right battery is available now or it will be soon.

Don't underestimate the brutal efficiency difference of wheeled vs quadrupedal locomotion.

The noise level was a pretty obviously non-negotiable flaw in the mule from early on. You'd think that someone would have pointed that out before dumping $42M into it.

I wonder if there were plans to reduce the noise in the unit that never worked out.

I think it was more for the practical research rather than actual use. Human response, movement tests, acceptable noise levels etc. All of that is very valuable in of itself.

There are electric versions by Boston Dynamics but they carry less weight which probably disqualifies them from the project. I'm sure someone was hoping that the firms involved would pump out an all electric version that could meet everything but it just wasn't feasible within the parameters.

I remember in one of the Ghost Recon Advanced Warfare games, there is a mini cooper sized APC looking thing that would carry supplies. It would either follow the player or could be controlled. I think that would be a better solution. A gas electric hybrid offroad vehicle would be pretty cool. It could carry supplies or serve as an automated medevac and still be silent in electric mode.

I went to a few demonstrations of the robot at an army base back in 2012. Whenever they would start the robot up, all the spectators would involuntarily gasp and cover their ears. It was shockingly loud.

Human-pulled wagon [1] with smart robot suspension?

may have Dalek [2] limitations ...

[1]https://en.wikipedia.org/wiki/Radio_Flyer

[2]https://en.wikipedia.org/wiki/Dalek

Looks like another case of 'we solved the wrong problem'. Too bad they spent all this money without asking anyone, 'hey, is it a problem if this thing is super noisy?' I think the takeaway lesson here is, if you are building something for the military, have an end user - a soldier - involved in the design early on, and listen to what that person has to say. That applies to many different kinds of engineering. Don't solve the wrong problem.

I feel bad for the individuals who invested their time into this project, I've been reading about it for years. It's like getting to the Olympics after a lifetime of training and when the gun goes off you trip and fall on your face.

Why not a hybrid drive train like the volt. So you run the gas engine to recharge in safe areas and run on batteries in quiet areas?

Also moderns cars seem extremely quiet. Why can't they apply similar silencing here?

It only has 400lbs carrying capacity. Batteries to give it a reasonable distance -- maybe a kilometer over rough country? -- would eat into that.

i think it can be used for police operations instead of dogs in SWAT situations like this for example http://www.dailymail.co.uk/news/article-3323495/Hero-police-...

There is no technological equivalent for a dog's nose in reliability, form factor, and cost. That, and a Boston Dynamics type robot probably costs several orders of magnitude more than training a dog/trainer pair, and is much less flexible.

Amusing note: at least for ATF, all arson/explosives detection dogs are dropouts from the seeing eye dog training programs (too energetic). Source: relative was involved with the program.

Cost would probably be factor number 300,000 when it comes to the types of missions it's expected to see. Any other kind of mission, and the modern military supply train is orders of magnitude more efficient.

>There is no technological equivalent for a dog's nose in reliability, form factor, and cost.

i specifically mentioned SWAT missions. This is where a robot can replace the dog or man "on the point". It just have to enter the room/etc. and burn several flash grenades, even without throwing them :) After that real people&dogs may enter much more safely.

$42 million and they couldn't design a quieter engine?

Why didn't they look at fuel cells?

The need is still there ...

1) American recruits are too fat to carry a pack nowadays

2) Marines in Afghanistan suffered from ankle injuries from carrying 70 pound packs on cobblestone fill.

Reply on news.ycombinator.com