Insects have long inspired humans, even before the dawn of time. There is a theory that ants inspired human agriculture. As any observant child will know, ants routinely take seeds into their nests, and occasionally, they will abandon those locations. Could they have inspired prehistoric man to carry seeds, place them under dirt, and care for them?
It’s an interesting theory, but unfortunately, one for which we’ll never know the answer. However, we don’t need to theorize or look so far back into history. In this article, I have five examples of how ants and other insects have inspired current technologies.
Five examples of Insect Tech: 🐜📡
- Ant Colony Optimization (ACO)
- Swarm Intelligence in distributed systems
- Mechanical Biomimicry
- Efficient Neural Networks by bees.
- Beetle Moisture farming.
- And a bonus example: Scroll down to find out!
Ant Colony Optimization Algorithms
Ant Colony Optimization (or ACO) is a collection of algorithms designed to find a good solution through networks and graph problems. These are some of the first examples of probabilistic problem solving. In 1992, Marco Dorigo first proposed ACO for his PhD thesis, and it was the 1st algorithm to search for an optimal path in a graph. His algorithm was modeled as ants seeking a path between their colony and food.
This idea was later diversified to solve many more problems, each drawing on the behaviors of ants. In 2004, scientists showed that ACO-type algorithms are closely related to stochastic gradient descent, cross-entropy method, and estimation of distribution algorithms. Gradient descent and cross-entropy are several of the main methods that AIs use to converge on a solution. So, ants’ strategies are proven to be similarly effective as AI.
ACOs have also bridged the gap between academia and industry. Theory and practice. The industry has used them on any NP-Hard problem in the categories of routing, scheduling, or assignment of resources. For example:
- Network routing (both connected and connectionless).
- Modeling protein interactions and protein folding.
- Optimization of Business Processes.
- Circuit designs.
- Routing and Scheduling: Airlines have also used ant-based routing in assigning aircraft arrivals to airport gates.
- Edge Detection.
The lowly ant 🐜, a symbol of industriousness, has proven to be more than a symbol. Industry has followed.
A great overview of ACOs is available here.
Swarm Intelligence (SI)
Move over, Sports Illustrated or the metric system (International System of Units (SI)), there’s a new acronym taking over. Swarm Intelligence is defined as a collective behavior of decentralized, self-organized systems, natural or artificial. These systems have been used since the nineties for computer graphics, embedding themselves in such movies as The Lion King or The Lord of the Rings Trilogy for scenes with many simulated participants.
I am going to ignore these applications, though they are directly influenced by insect behaviors.
Instead, I’m going to highlight how Swarm Intelligence is helping humans by aiding medical diagnoses. For example, swarms of human radiologists connected via SI software showed substantial improvements versus single radiologists or traditional machine learning examples (study).
Another study using doctors and MRIs showed substantial gains compared to a majority voting method. Each doctor was given a virtual magnet to ‘pull’ the collective output in a direction, like in the above image. The Doctor’s “hive mind” consistently beat single doctors or even AI (study).
Experts networked together exceeded the performances of a single expert AND artificial intelligence!
Mechanical Biomimicry
Ants are renowned for their strength-to-weight ratios. In this study, scientists tested the neck portion of ants. Using electron microscopes and micro-computed tomography, they measured the neck exoskeleton and the way it blended and combined into softer parts.
After first estimating a conservative 1000 times the body weight, they tested these sections and found an astonishing 3500 to 5000 strength-to-weight ratio.
“The interface between the soft and hard material of the head. Such transitions usually create large stress concentrations, but ants have a graded and gradual transition between materials that gives enhanced performance.” These sections contained blends of soft and hard material in such a way that gave it considerably more strength than estimated. This can lead to micro-sized robots which also combine hard and soft parts without any weak points.
Original Research.
Efficient Neural Networks
Bees are currently influencing the development by super-efficient Neural Networks. *Bees’ secret to super-efficient learning could transform AI and robotics * Now, ignore the “could” in this headline. This is one of those headlines that ruin the message by adding in the weak weasel word ‘could.’ ‘Could’ doesn’t mean it ‘doesn’t.’ Let’s focus on the reality of that sensational headline.
Right now, biologists and data scientists ARE deconstructing how insect neural networks ARE so efficient. It’s like an old, misleading headline about how bumble bees shouldn’t be able to fly, according to aeronautical engineers. Clearly, bumblebees have flown for untold numbers of years, and science has yet to understand how. Similar to their wings being a mystery, the way bees navigate 3d space with so few neurons is something science will unlock and enable a technological revolution in efficiency.
Scientists are adopting a ‘neuromorphic model of active vision’ and molding their models of neural networks to the efficient realities of nature. These should usher in lower-power capabilities for many existing technologies that rely on vision.
Original Study.
Beetle Moisture Farming
Time for the beetles to step up. Whether it’s recycling dung or making rocket fuel, the beetles have revolutionized more than music. Scientists have studied Namib Desert beetles to discover the secrets of how they have survived with little water. The secret is not solely in avoiding the hottest time of the day, but in the way the tiny organism’s wings alternate between structures.
Yes, by alternating between hydrophilic bumps with hydrophobic channels, the tiny beetle manages to force tiny droplets of water in the air into higher concentrations until the combined weight drops them into the mouth of the beetle.
Scientists have taken notice by doing the same alternation of structures to enable moisture farming (yes, that age-old sci-fi profession). Not only that, they can tweak your mirrors to prevent the only downside to a hot shower: fog. Fog-Free mirrors, windshields, glasses, and facemasks incoming!
Researchers at the Massachusetts Institute of Technology have emulated this capability by creating a textured surface that combines alternating hydrophobic and hydrophilic materials. Potential uses include extracting moisture from the air and creating fog-free windows and mirrors. A company called NBD Nano is attempting to commercialize the technology.
Now, those are some serious engineering insights from a tiny desert insect. There are even other beetles in the same region with different moisture acquisition strategies. Some build mounds in the sand to enhance fog droplets. Still others collect it off of silk or wind-blow detritus. Needless to say, those are also being studied for any technological insights.
BONUS! Insect Cyberware. Welcome to the Stigmergic society
In the previous examples, we’ve taken technology from the bugs; now, we’re putting technology into the bugs. I discovered this when doing research for my sci-fi novella Space Ants: Never Say Die.
It’s called Insect Machine Interfaces. And yes, these are real. Scroll back up if you want to sleep tonight. Insect-Machine-Hybrids are being researched and developed. In the above 2009 study, scientists and engineers discovered the best stage of the development cycle to insert electrodes into developing pupae in order to best secure a connection between the nervous system of a developing insect and inserted microelectronics.
The early pupae stage provided the best mechanical and electrical coupling.
Not to be outdone in weird science, this other study from 2012 is titled: Insect-machine Hybrid System: Remote Radio Control of a Freely Flying Beetle.
That study also links to additional research where scientists do the same things with walking insects, such a spiders.
It remains to be seen whether these abominations configurations are actually more effective than microdrones or other fully robotic situations.
It seems hard to mass-produce such insect cyborgs, not that they are not stealthy, though I suppose someone could try and hide the microelectronics.
However, will this usher in some weird hybrid society, the Stigmergy Society of AI and humans, and insects collaborating?
Everyone was worried about AGI, or Artificial General Intelligence, taking over; they never thought to look for the other AGI: Artificially Guided Insects.
