You have to work out what this can do that other stuff can't due...

You have to work out what this can do that other stuff can't due to danger cost for military or industry etc etc..

The sensor pack is the key here and for example if a millitary wanted to do a survey or a beach at high tide that they were going assult then with the correct package they could get a very current picture of the surface they would be landing on in the shallows etc. The key to this is would this survey be detected by countermeasure sensors if they were in place etc etc.

The key is if the benefit of getting all the data at one in a short time from multiple units with built in redundancy and only if it is possible to have sensors small enough to be carried etc etc

As for putting explosives in them? Need to think od a suitable situation where there is a advantage of multiple smaller units and explosions over a larger torpedo, mine surface to air or conventional shell and that may be as simple as small units can get through defences and combine at point to create the larger explosive but that isn't really correct..


The distributed data collection also means that there is redundancy as if a unit dies them program can reset others to fill the void it leaves or at least all data isn't lost just that one data point or points one of the swarm could have provided so in places where there is a small window of "opporunity" such as a large tidal change area you by having multiple units - swarm collect the full data set very quickly and at lower risk all assuming you can afford to provide x number of sensors economically. .
In survey work they could be used as portable sensors to collect data in reagardsto 3d explosive surveys in shallow water?

Tell me a killer application as I have had a think and not come to a no brainer example yet .

It will get the boffing thinking as to what tasks previously have been too hard or machine , cost , labour intensive that could now be completed with smaller more portable units.