[scroll to the end for latest post]
I am intrigued by the flocking behavior of birds and schooling fish and the swarming behavior of insects like bees. Here are a couple of videos that clearly demonstrate swarming or colony behavior…
- Incredible Starling Murmuration: https://youtu.be/L7CyMaOAqyo
- Ten Million Starlings Swarm: https://youtu.be/UVko9jyAkQg
- Hawk attacks Starling Murmuration / Swarm over fields: https://youtu.be/_btcPA7ssgc
- Amazing Fish Form Giant Ball to Scare Predators | Blue Planet | BBC Earth: https://youtu.be/15B8qN9dre4
- Swarm Of Locusts DEVOUR Everything In Their Path: https://youtu.be/6bx5JUGVahk
- Ants Battle to Protect Their Fortress: https://youtu.be/qu0HN9rYtIw
I think it is both beautiful and fascinating as one observes a flock of birds flying together in different directions within a larger group, each without touching the others. Any time that I observe one of these enigmas I ask myself several questions.
- How could I film this behavior without losing the 3D forms that are exhibited as the flock moves and changes directions.
- What motivates the flock?
- Is there a lead bird (like in the migrating “V” behavior of Canada geese) or does each member have the same status in the flock?
- How does each bird sense its neighbor in all directions so that it does not accidentally fly into another bird (creating a domino effect which could bring down several birds).
- Are there rules that govern this behavior? How are they implemented or enforced? When happens if one or more bird ceases to follow the rules–would there be a cascade failure of the flocking behavior?
- Is there elasticity within the relationships such that each neighbor bird moves away from the errant bird (if it happens that one bird does not stay within the set of “rules” for flocking behavior).
- Can I recreate this behavior mathematically and display it in a 2D space like on an HTML5 canvas?
- Is the flocking behavior random or purposeful?
- Are flocks homogeneous (same species) or are there mixed species flocks or swarms?
- Are the swarming behavior in insects, flocking in birds, herding in mammals, and schooling in fish all related by the same underlying system or mechanism? Can the same mathematical model be used to represent them all the same way?
- What creates the illusion of fluidity as demonstrated when a predator enters a school of fish and they move together but away from the intruder?
- Is this behavior relevant to autonomous or partially autonomous AI drones?
- If this behavior can be programmed into swarming drones could this have positive implications in military applications like simultaneously clearing a building and then moving on to the next while one keeps watch on the cleared building while the others continue to examine the next buildings for people and weaponry? My thought is that this could save lives.
- Could the drones be armed but still small enough to swarm together through windows, doors, and other openings? How might a drone swarm be used in peaceful missions (medical emergencies, search and rescue, etc.).
- How might a swarm of drones be neutralized and how could one protect against this neutralization? (Military implications and countering terrorism). Are drone swarms susceptible to hacking in real time?
- What ethical and privacy issues would need to be addressed?
- Could swarming behavior apply to some sort of nanobots that would enter a system and help to heal an individual? How would they be removed or eradicated/dissolved when their job was done?
- Can drone swarms contain individual communication and “intelligence” that act together as well as connection to a centralized computer (to help isolate the system against a single point of failure or hacking). Could these drones share their intelligence and mission with each other to reset any remote programming (hacking) interference? Can the drone swarm act as an intelligent organism based on inputs from the individual members? What factors would govern its purpose or mission if it was dynamic? (I think that this would be analogous to slime molds in nature where individual units act together as an intelligent mobile organism with built-in error correction and reorganization if the whole is split up). This article is interesting itself as a standalone topic…”https://www.discovermagazine.com/planet-earth/what-is-slime-mold”.
Here are some references for review:
Animal flocking behavior:
- https://birdfact.com/bird-behavior/social-structures/flocking-behavior
- https://www.semanticscholar.org/paper/The-chorus-line-hypothesis-of-manoeuvre-in-avian-Potts/64ed6452f1af3f43a7d4882b8bc3e1708b735b0e
- https://arxiv.org/pdf/1705.01213.pdf
- https://sora.unm.edu/sites/default/files/journals/wilson/v100n01/p0108-p0118.pdf
- https://en.wikipedia.org/wiki/Swarm_behaviour
- https://en.wikipedia.org/wiki/Swarm_behaviour
Robotic drone swarming:
- _Good overview: https://link.springer.com/article/10.1007/s43154-021-00063-4
- _https://dronelife.com/2021/01/14/french-drone-maker-unleashes-autonomous-swarm-concept-at-ces-2021/
- _https://www.edrmagazine.eu/escribano-designs-a-swarm-system-of-uavs-for-surveillance-and-recognition-missions
- _https://www.youtube.com/watch?v=44KvHwRHb3A&ab_channel=GuinnessWorldRecords
- _https://www.adsadvance.co.uk/blue-bear-demos-collaborative-20-drone-swarm-on-bvlos-ops.html
- _https://www.cnn.com/2018/09/04/health/nano-swarm-robots-intl/index.html
The mathematics of swarms:
- https://math.umd.edu/~tadmor/pub/flocking+consensus/Tadmor_swarming_Notices2021.pdf
- https://www.math.uvic.ca/faculty/rillner/papers/swarm_august6_2010.pdf
- https://www.math.univ-toulouse.fr/~cnegules/Article/Drones.pdf
11/4/2023
- Recently I have been observing more flocking behavior in birds and insects as they prepare for the winter migration. This evening I took a walk with our daughter to enjoy the crisp autumn weather. We sat down on a wooden arbor bench in the middle of our flower garden and in the process disturbed a swarm of at least 35 gnats, each about twice the size of a large fruit fly. These insects were backgrounded by a bright cloudy overcast sky so we could see them clearly. I watched individual insects as well as pairs of them and I made the following observations:
- No single insect traveled more than 30 feet (or less) from the center of the swarm. So there must be an “elastic distance” similar to what we would observe if each was tethered to a long piece of elastic. One question I have is whether individual members of a flock have some relative knowledge of the position of all the others, or if perhaps some “rule” relating them to their nearest neighbors causes the appearance of an overall rule or motion for the entire flock.
- No two insects would come closer than 2 or 3 inches to another gnat in the group. I will call this the “bounce” or “rubber” distance between individuals.
- Each individual seems to prefer to maintain a more-or-less constant average distance from the other nearby members but if they move out of this “comfort zone” then the elastic and rubber rules seem to apply.
- The whole swarm seemed to hover randomly as a large ellipse then they moved off together as a group. I have recently seen birds do the same–the flock moves in seemingly random (but fluid) directions until it settles on a main direction and then the whole flock moves in that direction.
11/9/23…Additional thoughts:
Each member may respond individually to some stimulus or motivator like food (moving toward) or danger (moving away), or perhaps only the “shell” members respond (those near the periphery of the flock who can see or hear or smell external stimuli better than those members within the core of the flock). Or there could be some sort of independent “self will” by individual members (personal preference to move in a certain direction at a given moment). Assuming the individual members of the flock or swarm continue to adhere to the rules of the flock, this could help explain why parts of the flock appear to move in a direction away from the main forward direction of the flock then merge back in without breaking off from the main flock as the rules of the flock reimpose on the “wayward” members and cascade like a domino effect throughout the flock in all directions.
Based on my preliminary observations this would mean that the position “P” of a given individual “I” at a given time “T” could be described as a function “f” of the individual’s initial position “p”, the group elasticity “e”, the rubber or bounce factor ” b”, a self will factor “w”, and a stimulus response factor (positive or negative) “s”. The position of the flock as a whole would then be described as the summation from 1 to n of the individual functions “f” where “n” is the total number of individuals in the flock at time “t”.
11/13/23…Small flocks.
Two new observations are in the evening for small flocks of birds with 5 members or less. First, in the evening the elastic distance may be just the distance to the nearest treetop as the birds perch for the night. Another observation is that with small flocks the birds fly more closely together if they are mated pairs. I am not sure if this applies to flocking behavior in large flocks, but I am guessing that it does have an effect on both the elastic distance and the bounce factor.
02/27/24…update on DHS/military anti-drone tech:
Apparently there have been some problems with some civilians flying drones illegally in restricted airspace and even shutting down airports. This technology may help to elimanate these situations and help apprehend the perpetrators. Also there is an interesting mention of “dark” drones that use a different RF technology. This anti-drone technology of course begs the question of how to counter it in military applications and swarms. https://www.dhs.gov/science-and-technology/news/2024/02/27/feature-article-st-tests-cutting-edge-counter-drone-technology .
07/29/24…update on Drone ideas:
Given the recent assassination attempt of the Republican nominee and the fact that drones were apparently used by criminals to scope out the area, I propose some additional ideas to my drone swarms. In addition to autonomous and collective intelligence in the “web” of drones in the swarm there should be at least two stealth drones that remain hidden. The purpose of these is to detect any incoming anti-drone defenses such as bullets/nets to physically disable them. The surveillance drones should operate on a different encrypted set of wavelengths (perhaps ones that are unlikely to be blocked such as cell tower frequencies). The surveillance drones would be analagous to AWACS and relay info to the drone swarm. (I am stating this in the context of a battle zone, not for use in public surveillance). In the case of the recent assassination attempt a couple of weeks ago, such a drone swarm could have been deployed hours in advance in lieu of the dubiously stated reason of lack of staffing to protect the high profile candidate. To address the power/battery issues of the drones, there are portable “land and recharge” platforms that could be deployed at specific locations on the site, such that half of the swarm would still be in the air at any given moment over the course of the surveillance period. Another option is to add gas or other fuel-powered drones into the mix. Ideally, sensory information to be included would be attempts by other sources to launch and deploy drones other than those authorized and registered at the site. Depending on the public law, any legal non military drones should have to register temporarily within a certain time period of the event with the understanding that any and all drones may be grounded or shot down as necessary to protect the prominent public figure. Fuel powered drones could also include intercept and block (or destroy) capability to take down UAFs.