In the modern-day military use of dogs in combat is fairly limited; the United States Armed Forces has about 2,500 "working dogs" serving, compared to the 1.2 million active humans (1:480).

The goal is to increase the number of "working dogs" drastically (>= 1:50) by changing factors relating to the animals and their roles.

Assuming that we can handwave some impressive biological technology to create a superspecies, and be very liberal with the term "working dogs," assume there exists technology to create a creature with combination of traits (and only the traits) known to exist in modern or historic canines, vulpines, or felines.

Could the intended ratio be achieved? Would this significantly alter warfare on any level (logistics to field operations)?

  • 2
    $\begingroup$ Those aren't your typical guard dogs so breeding that many is going to take a toll on the defense budget because only a fraction pass the amplitude test is considered war asset, what about the remaining animal? Of course GMO pooches can pack a punch on the battlefield but what happens if they escaped or when their owner is kia? They perform better through 1 on 1 bonding and expect them to turn against each other from time to time. $\endgroup$
    – user6760
    Commented Oct 19, 2016 at 17:04
  • $\begingroup$ Something you might want to keep in mind is that the geneva conventions don't apply to animals, so there is no reason to not use chemical or biological weapons against your dog army. Like a weaponized disease which only affects canines. $\endgroup$
    – Philipp
    Commented Oct 20, 2016 at 11:49
  • 1
    $\begingroup$ @Philipp - True, the Conventions don't. However, see the Hague Conventions (1907), Geneva Protocol (1928), and the subsequent Biological Weapons Convention (1975), and CWC (1993) which prohibit making/stockpiling those weapons, regardless of species. (Although a lawyer could probably weasel around that). Plenty of reason not to create/use canine targeted bioweapons. $\endgroup$
    – JohnP
    Commented Oct 20, 2016 at 14:48
  • $\begingroup$ I say work backwards, starting with the many existing types of dogs and then finding a role for each type. To encourage creative dog employment, there could be penalties for units which did not achieve the >1:50 ratio. $\endgroup$
    – Willk
    Commented Jan 10, 2017 at 18:59

6 Answers 6


Your ratio of dogs to personnel is a little misleading. It is relevant to note that there is pretty much no use for dogs in the Navy and Air Force. There are 460,000 + 182,000 = 642,000 active duty personnel in the Army and Marine Corps.

Furthermore, most of the Army and Marine Corps are flabby support personnel (dirty POGs). To get a semi-reasonable count of the units that can actual do fighting (not counting air units who don't have much use for dogs either) lets count up the Brigade Combat Teams for the Army. For the Marines it is more complicated because the relationship between MEF, MEB, and MEU is not fixed, but we will try to sum the strength of the Ground Combat Elements.

For the Army, there are 32 brigade combat teams at a strength of 4400 soldiers each for 140,800 actual soldiers. Estimating 73 battalions at 900 Marines in the 4 Marine Divisions combined, we get 65700 Marines. This means a total of 140,800 + 65,700 = 206,500 total actual combat troops. Taking your number of 2500 working dogs the ratio is more like 1:83; a lot closer to your target.

To close the gap, the only thing we need is a bit of culture change. From conversations with several brothers/college roommates who have been infantry officers in both Iraq and Afghanistan, everyone wants more dogs.

First off, Iraqis are generally terrified of dogs (see Abu Ghraib). So there is that.

Second off, when there are bad guys in a building, and you want them dead, but there are women and children in the mix, dogs are a great option. They are trained to go after people with guns, so they are much less likely to kill women and children (especially compared to, say, artillery).

Third, after I expressed incredulousness at the notion that dogs are so effective as manhunters, my friends informed me that guerilla fighters are generally unprepared to deal with a dog. They don't wear protective clothing (camo with tough fabric is a big help if you are being bitten), they don't have training in hand to hand (or hand to dog) combat, and since they are fighting soldiers, they tend to have guns instead of knives. Guns, however, are much less useful than knives in a small room with a 120 lb German Shepard hanging on to your arm.

In any case, consider that anecdotal, but if your army is primarily into anti-insurgency warfare dogs are a pretty good deal, no genetic engineering required. If the generals actually listened to the JOs (or, heaven forbid, the NCOs) there would probably be more dogs around already.

  • 1
    $\begingroup$ Not sure about your Navy and Airforce comment. Wikipedia shows that the Air Force had about 1,700 dogs in the 70s, and currently has about 500 dogs, so the modern Air Force has (give or take) about 1/5 of the current "working dogs." On another note, you bring up some excellent considerations later on, so thank you for that. $\endgroup$
    – Ranger
    Commented Oct 19, 2016 at 15:33
  • 2
    $\begingroup$ @NexTerren In the past they used the sentry dogs to keep the Viet Cong from getting into their bases. Present day Air Force dogs are law enforcement, probably for drug sniffing. I can say the same with more certainty about Navy working dogs. $\endgroup$
    – kingledion
    Commented Oct 19, 2016 at 15:38
  • $\begingroup$ Dogs would be very useful to Air Force and Navy since they do certain army-like things: site security for instance. Coasties can use drug-sniffing dogs. The Navy SeaBees do battlefield construction and could sure use a 15 lb. mine-sniffing dog. $\endgroup$ Commented Oct 20, 2016 at 4:07
  • $\begingroup$ I thought it was a principle of the Marine Corps that everyone is a rifleman, no? $\endgroup$
    – Dronz
    Commented Oct 20, 2016 at 5:34
  • $\begingroup$ @Dronz Of course, Devil Dog, OOOOORAAAAAAAHH!! But seriously, ask a 5954 when the last time they went to the range was. $\endgroup$
    – kingledion
    Commented Oct 20, 2016 at 12:10

Your biggest limiting factor is actually going to be training/handlers.

For example anti-tank dogs were a thing! But because the training fell short, in practice they ended up getting scared and running back to the trenches, blowing up its allies instead of enemy tanks (though, I kinda feel they had this backfire coming for even attempting it)

Out of the first group of 30 dogs, only four managed to detonate their bombs near the German tanks, inflicting an unknown amount of damage. Six exploded upon returning to the Soviet trenches, killing and injuring soldiers.[3] Three dogs were shot by German troops and taken away, despite furious attempts by the Soviets to prevent this, which provided examples of the detonation mechanism to the Germans.

You will need a reliable way of training these dogs, make sure they remain loyal/on duty during a mission, and remain coordinated with their handlers (which will be the hard part. You either need the dog to be reliably autonomous or restrict their actions to remaining near the handler.) In war, the more variables you introduce, the more likely a mission is to fail. So you really just need these dogs to be about as reliable as the soldiers they work with. (depending on their reliability, they can either make or break your army. like the anti-tank dogs)

  • 4
    $\begingroup$ A well trained dog is too expensive to sacrifice. You need a magnetic grapple on the satchel charge, so the charge grabs onto the tank, and releases from the dog; this tells the dog to run home. The computer on the charge is sacrificed instead of the dog. $\endgroup$ Commented Oct 20, 2016 at 4:00
  • 2
    $\begingroup$ For delivering explosives, we have stuff like this already. It's called UAV drones. $\endgroup$
    – Nelson
    Commented Oct 20, 2016 at 8:07
  • $\begingroup$ Although, really my main point about the anti-tank dogs is that, even with training, they are unreliable in a war zone. And if you read the wiki, it sounds like the handlers were displeased with having to kill their own dogs (or else get accidentally blown up by them), and refused to take on more. So there is a sympathy aspect with how much danger you can reasonably put the dogs in. $\endgroup$
    – Tezra
    Commented Oct 20, 2016 at 12:34

Why dogs?

  • They scare unarmed people. Dogs are useful for crowd control and prisoner control.
  • They assist scouts.
  • They can be trained to detect explosives, drugs, etc.
  • They guard installations.

If you want serious growth in numbers without the "handler bottleneck" identified by Tezra, get more installation guards. One private can feed dozens of dogs in a kennel. And if they can roam freely in the area between two fences, there is no need for trained handlers.

But that is probably not what you wanted. The goal would be to get one dog into each rifle platoon or rifle squad ...

  • Find a breed that will fit onto a mostly-human "pack" without fuss and accept a place at the bottom of the pecking order. It must accept that a soldier fresh from the replacement depot outranks the most senior dog.
  • It must be clever enough to accept orders like "attack anybody in that house" or "find out if there is anybody in that house, then report back" and to communicate the result to any squad member.

A dog that can work without dedicated handler.


Historical Examples

Note that there are some tales of dogs of their own accord helping soldiers in war. One case, a dog got angry and ran into the enemy German's machine gun nest. It was actually shot by the machine gun a few times, too. The Germans ran out and surrendered as soon as the dog got in (which indicates they didn't put up much of a fight). You also of course had the many rescue dogs who looked for wounded soldiers and helped them back to camp.

The other major historical example of a war dog is to chase down fleeing enemies, in order to attack them from behind and inflict casualties.

Armoured Attack Dogs

You could start armouring the dogs like the samurai did for their hunts, then try to breed some variety of dog so aggressive it will attack the enemy even when artillery and gunfire is going on (and hope it doesn't use that aggression against you...). Dogs would not be bad trench clearing companions, if they would stick close to you and perform under such high stress, many do not have that gumption.


Realistically, you would not only a lot of effort put into the breed, but then to select the dogs that demonstrate the necessary loyalty and bravery and aggression. It would be a pretty questionable investment, compared to spending that on equipment and training of more men. The real advantage to dogs is they're cheap, so making them expensive is counter intuitive.

Cheap Strategy, Animal Farms

A cheap strategy would just be to take an aggressive breed of dog, or leopard or whichever, and raise whole families of them on your military bases and defensive lines. When intruders come around and cause trouble, the animals (if you select highly territorial examples) may instinctively start sneaking up on the enemy and killing them any chance they get, causing attrition (leopards are scary good at that).

Machine-Gun Dogs

You could consider higher tech methods, like adding machine-gun drones to the backs of dogs. One of the issues we have with drones is navigation of the environment, a dog smart enough to go near where the enemy is could give the drone a line of fire. That does come with the issues of jamming, miscommunication, mistaken identity, and EMPs that drones come with, of course. Plus the issue of a dog running around with a machine gun which may just glitch out and start killing civilians or friendlies.

  • $\begingroup$ How about suicide bomber dogs? $\endgroup$
    – Chris H
    Commented Oct 20, 2016 at 9:22
  • $\begingroup$ @ChrisH, Bomber dogs might work as a last-ditch effort, but probably not as a proper military strategy. Even your link mentions a lot of failures and that "donkeys <...> were more reliable" $\endgroup$
    – user8808
    Commented Oct 20, 2016 at 11:41
  • $\begingroup$ @Roux, true enough. The WWII Russians claimed some successes in a desperate (last-ditch?) situation. But they are a real example in modern-ish (machine guns, mechanisation, etc.) warfare. $\endgroup$
    – Chris H
    Commented Oct 20, 2016 at 12:42
  • $\begingroup$ Another answerer actually went into the suicide dogs, so I thought I shouldn't cover that a second time. We're probably getting close to the time where little remote control drones will be better for that. $\endgroup$
    – J. Doe
    Commented Oct 20, 2016 at 16:04

In Heinlein's Starship Troopers, the Mobile Infantry had neodogs, which were a biological uplift "superspecies" of military working dog with the ability to speak human languages (more or less) and a much increased intelligence. (Their breed name was 'Caleb'.)

A neodog was typically paired with a human partner and the two shared a deep psychological (possibly even psychic?) bond that enabled the two to work extremely well together, but made the loss of one partner almost impossible to bear for the survivor. Anyway, they're not mentioned all that much in the book, but they were essentially used as scouts, where the dog's superior senses and hardiness were most useful. I could easily see extensions of this role to include:

  1. LRRP (Long Range Reconnaissance Patrol): They might operate in packs without human support, or with one or two human partners to provide additional direction and supervision. Their biggest logistical advantages would be if they could operate without humans, since dogs would presumably be much better able to survive and perform at peak capacity in wilderness conditions, forage more easily (and require less food), move more stealthily, move more quickly without vehicular support, and so on. Roaming dog packs might also arouse less suspicion if seen by enemy troops.
  2. Special Hazard Duties: Depending on their modifications, these neodogs might also be trained for special duties that are extremely hazardous to humans, like demolition and mine/IED detection and disposal, navigating paths through minefields, clearing trenches and caves (as they were used in the island campaigns of WWII).
  3. Force Protection: Neodogs would be ideal sentinels, as they have superior hearing and sense of smell to humans and better night vision, and perhaps greater stamina and less likelihood of falling asleep or getting bored/distracted while on patrol.
  4. Combat Search and Rescue: Dogs are already used for this now, and a more intelligent dog could be presumed to be even better, following more complex orders and better able to detect and evade enemy patrols for, say, extraction of downed pilots or wounded soldiers caught behind enemy lines.
  5. Guerrilla Operations: Fast moving, harder-to-detect, vicious attack dogs prowling the forests and penetrating deep behind enemy supply lines to attack and kill at random might produce a demoralizing effect that is far outsize of the actual material damage they do.
  6. Couriers: There may be times when it is impossible or perilous to deliver messages or parcels via technical means (heavy jamming, enemy has broken your encryption, enemy air superiority or ambushes make convoys unlikely to survive). Since dogs are faster than a human on foot and more stealthy, and a much smaller and quieter target than vehicles, they might perform this role admirably even over extensive rough terrain.
  • $\begingroup$ 1 consider small (not easily noticed from a distance) video and communications equipment so a human could remotely see and issue commands (drones would still be better). - 2 and 4 already done but of limited effectiveness soon to be outclassed by drones - 3 drones are vastly superior here and dogs are very easily distracted - 5 this is a crime against humanity - 6 quite a massive resource expenditure for such a small niche and still outclassed by autonomous courier drones $\endgroup$ Commented Oct 21, 2016 at 18:47
  • 1
    $\begingroup$ The endurance of dogs, who could potentially live off the land for weeks or months, far exceed that of drones. They also are unaffected by hacking or a heavy ECM environment. Drones used in search and rescue are used in the search part. They don't actually rescue anybody: they call a human over to help. One of the large benefits of a neodog over other dogs and drones in these situations that you seem to be overlooking is that the neodogs could be entirely autonomous. $\endgroup$ Commented Oct 21, 2016 at 19:52
  • $\begingroup$ Engineering hyper-futuristic neodogs with indefatigable focus is beyond the scope of the question, but at that point the tech level would allow for autonomous AI with superior capability. Drones can process energy sources (organic and otherwise) more efficiently, plus wait states give them superior endurance (and shelf-life) . Dogs biting and dragging people out of rubble might count as rescue, but robots offer better adaptable capability there too. Comparing 'neodogs' against drones frozen forever at current consumer-grade tech demands more suspension of disbelief than I can muster. $\endgroup$ Commented Oct 27, 2016 at 19:32
  • $\begingroup$ Beyond the scope of the question? From the OP: "Assuming that we can handwave some impressive biological technology to create a superspecies..." So it's not beyond the scope of the question; rather, it's assumed in the question. Comparison of fictional entities is inherently speculative, so we can just as easily assume that neo-dogs are more effective than your hypothetical drones. The question asked about how to make uplifted canines effective, not how to ignore them in favor of drones. $\endgroup$ Commented Oct 27, 2016 at 19:45
  • $\begingroup$ the question states "create a creature with combination of traits (and only the traits) known to exist" - can you show me a modern dog or cat with hyper-intelligence? With the ability to exhibit long-term concentration free from distraction? for complete autonomy for a period of months while still adhering to complex conditional orders? Even picking individual traits from a wide variety, you are presuming traits which do not (and probably cannot) exist. $\endgroup$ Commented Oct 28, 2016 at 17:57

I recall a co-worker telling me about his time in the (first) Gulf War. His duty was to guard a plane and he spent the shift walking around it. He recalled that when the temperature reached a certain point, the dogs would be taken in, as they can’t handle that heat.

That might be why (as related by kingledion) Iraqis are generally terrified of dogs: they are not common there.

So a significant factor in using dogs would be to have a theater of operations in a region where the climate was suitable for dogs to live.

  • 1
    $\begingroup$ There are breeds of dog that cope well with heat. You could for example use the canaan dog, which is adapted for a desert climate, as a starting point. $\endgroup$
    – user
    Commented Oct 21, 2016 at 19:11
  • $\begingroup$ There must be some reason why that wasn’t done. $\endgroup$
    – JDługosz
    Commented Oct 21, 2016 at 19:21

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .