Swarm Robotics Breakthrough Brings Pheromone Communication To AI (thestack.com)

An anonymous reader writes: Computer scientists at the University of Lincoln have invented a reliable, low-cost system which replicates in robots the pheromone-based communication behind insect swarms. Using off-the-shelf equipment including an LCD screen and a USB camera, the team has proposed what they call COS-phi, or Communication System via Pheromone. The artificial pheromone trails are traced visually onto the screen. As soon as a bot picks up on the path, it is forced to follow the leader.

Comments

Forced to follow...

[JustAnotherOldGuy]

"Forced to follow..."

So it's like the behavior of Millennials when a new iPhone comes out?

Re:Forced to follow...

[Anonymous Coward]

No, not exactly. There are some differences between these robots and Millennials.

Unlike Millennials, these robots don't force "tolerance" by any means necessary, including intolerance that far exceeds anything the allegedly "intolerant" person may have engaged in. (See the case of Brendan Eich.)

Likewise, unlike Millennials, these robots aren't totally into the complete and ruthless censorship of anything deemed to be "politically incorrect".

And again, unlike Millennials, these robots don't claim to support a million different social causes, none of which they really care about, just to make it look like they do care.

In summary, these robots tend to be better, more well-rounded, reasonable people than Millennials are.

This is a breakthrough?

[thesupraman]

Interesting to see what people consider a robots breakthrough these days.

It seems little more than the old follow the white line 'robots' (that have existed for for at least 50+ years, and primary school kids build with a couple of photodiodes, motors, etc) a monitor for them to run on (and not even a very large one), and a simple feedback camera to 'draw' from one of the robots.
They could do pretty much the same thing by tying a pen to the first damn 'bot' and putting them on a normal floor.

This has exactly ZERO do to with pheromones, swarms, or I would suggest breakthroughs.

I would expect something like this from a high school science project, not a damn PhD.

Hell, whats wrong with some actual chemical sensors, and droplet sprayers on the robots? would be more interesting and allow for different
mixing, trail lifespan testing, etc.
Or they could have used a virtual trail system using radio location and swam communication, but that may have involved actual development.

I am not sure which is sadder, this 'research', that a university actually allowed it, or that slashdot reported on it.

Re:This is a breakthrough?

[JMZero]

Yeah... I really don't understand a lot of robotics research. They seem to be forever chasing these awkward "proof-of-concept" implementations of concepts that are completely uninteresting. This is a perfect example: obviously you could make some robots that could do this, but it's really unclear what you'd learn by doing so, and the result is useless.

I mean, if they actually wanted this behavior for some purpose, and this was a reasonable way to approach that practical purpose? Sure, do it. Of course. If there was some question whether this was possible (which obviously there wasn't, at least not serious), then sure, prove it's possible. If there was some question we'd be answering (about biology, maybe?) or some challenge that would be interesting to overcome, then, uh, maybe. But it seems much more likely that such questions could be answered easier by a simulation.

As it stands, it seems like we've got a HUGE surplus of research on, to pick a random example, how to have robots all go somewhere without bumping into each other - most of it saddled with arbitrary restrictions that are also completely uninteresting. Can the swarming robots co-ordinate without any predetermined communication protocol? Yes, obviously, and we can prove that in simulation. But that's not good enough for some reason, have to actually build some crappy robots to mong into each other while we re-solve a bunch of boring practical problems with batteries and light sensors.

If you just want to have a challenge, or get some practice building robots that have to deal with the real world, why not build them to at least attempt something interesting or useful, or at least entertaining?

Re:This is a breakthrough?

[Anonymous Coward]

This has a lot to do with pheremone trails, like ants use. But since we can't replicate the chemical nature of pheremnes yet, they are sing photons. If you watched the video you could see that the trail fades with time as would pheremones and it can be followed the wrong way. They also showed following the stronger trail if it crossed itself. These are all behaviors that we had no way of duplicating until someone realized that a light trail could provide a useful analog that could be modulated (color, patterns, etc) to provide different "pheremonal" instructions. So yes, this research is far more than find and follow the white line, and it has quite a lot to do with pheremones.

Re:This is a breakthrough?

[Anonymous Coward]

This has a lot to do with pheremone trails, like ants use. But since we can't replicate the chemical nature of pheremnes yet, they are sing photons. If you watched the video you could see that the trail fades with time as would pheremones and it can be followed the wrong way. They also showed following the stronger trail if it crossed itself. These are all behaviors that we had no way of duplicating until someone realized that a light trail could provide a useful analog that could be modulated (color, patterns, etc) to provide different "pheremonal" instructions. So yes, this research is far more than find and follow the white line, and it has quite a lot to do with pheremones.

Except pheromones are biologic chemical signals to induce behavior in creatures of the same species, by definition. Using light that fades may be analogous to how pheromones dissipate over time but it is NOT pheromones! Pheromones induce the foraging behavior, not just provide a path to follow. The whole use of the term pheromones related to this research is misplaced at best, and misleading (fraud) at worst.

Re:This is a breakthrough?

[oic0]

If I had to guess, I would say they need to make some physical robot and make it do something they can relate to a biological behavior to keep getting money. Simulations and papers are good and all but you need physical demonstration and buzz words to woo and keep on riding the train.

Re:This is a breakthrough?

[lorinc]

Also, it has been done for many years. You can date the stigmergy based robotics back to the mid-nineties (cf the work of Alexis Drogoul for example) and the mathematical modelling dates back to the eighties (Deneubourg et al.).

What is presented here is nice engineering stuff, but has nothing to do with research breakthrough.

Re:This is a breakthrough?

[drinkypoo]

Yeah... I really don't understand a lot of robotics research. They seem to be forever chasing these awkward "proof-of-concept" implementations of concepts that are completely uninteresting. This is a perfect example: obviously you could make some robots that could do this, but it's really unclear what you'd learn by doing so, and the result is useless.

You get to focus on the software operating in an ideal environment, so you get to explore strategies before having to deal with the complexities of making the actual sensors work. There's nothing wrong with the idea, the problem is when you stop there. A lot of this work is done at universities by students, though, and they're learning. They don't know what they're doing. Their primary goal is to learn how to make the software work, not how to get good data out of a specific sensor that might not even be on the market by the time they get out of school.

So...

[BlueStrat]

So, if we can perfect small and sensitive pheromone sensors and install them on drone swarms, with a sample of an individual's pheromones they could be made to follow/target said individual, like for instance Putin or Obama.

Interesting.

I wonder if we'll start seeing FSB and SS agents following along behind said persons wherever they go, spraying cleaner/deodorizer or some other sort of pheromone-scrubbing chemicals and confiscating anything they touch in order to attempt to hide or obfuscate the unique pheromone trail these individuals leave wherever they go?

Strat

Re:So...

[nicoleb_x]

Countermeasures are trivial.

PacMan Smell

[MacroSlopp]

Sounds like they invented PacMan smell. (30+ years later)

And here I was

[YoungManKlaus]

thinking humans already had pheromone communication ...

Breakthrough?

[jandersen]

Is it just a bit rich to call this a breakthrough? I mean, it is a clever enough idea, taken from an already well understood phenomenon; but a breakthrough would be when you have worked on a hard to solve problem and then finally find a solution. Examples: Einstein's GR, Darwin's theory of evolution etc.

Hah, what?

[wonkey_monkey]

Other swarm robotics research has looked at alternative resources such as alcohol, light and sound to simulate pheromones and swarm behaviour. However, these are complex and expensive methods compared to COS-phi, which simply combines the LCD screen and USB camera with an open-hardware micro-robot and an open source localisation system.

Oh, yeah, sure, those are other methods are really complex. All this this needs is to constrain the robots to roaming around on top of an LCD screen. That's practical.

Just because nature does something one way, doesn't mean it's the best way.

Re:You've got to be $#!77!N9 me

[Hognoxious]

The summary and the article use "artificial" and "simulated" interchangeably.

Am I the only one who thinks they aren't the same thing at all?

Seems to me they might as well model it mathematically and cut out the wooden table stage.

Great tech, read Kill Decision for more fun

[virtcert]

Fabulous. We now have the missing technology for the pheromone-based killer drone swarms featured in Daniel Suarez's Kill Decision

RTFM

[Guyle]

Fabulous. We now have the missing technology for the pheromone-based killer drone swarms featured in Daniel Suarez's Kill Decision

Seriously someone should have read the manual before actually BUILDING them.