Biomimicry — Evolution Of Artificial Intelligence

10 mins read

Artificial Intelligence has taken the world by storm. Every news piece, geeky gossip or even people who aren’t so much geeky you would find them talking about artificial intelligence. Artificial Intelligence has evolved from being a technology once used by the military, software research labs to mainstream in the consumer electronics market. Especially with products like Robot Dogs, Tesla, Siri, Alexa, Google assistant and few other Google AI suites of products launched a sometime back. Also at the #WWDC18 apple has made it clear their platform (Mac OS) is going to enable developers to build AI applications.

In this article, we’d look at what is Biomimicry? How did artificial intelligence evolve as a result of Biomimicry? This article sets up the stage for a series of articles to follow on “Nuts and bolts of Artificial Intelligence” will be published later. In which we will deep dive technically with some code examples. However, in this article, we will limit ourselves to some fundamental concepts, case studies. If this sounds interesting to you, continue reading :-)

Humanity is dependent on Nature to survive, yet our society acts as if this is not so. Energy gained from nature powers our very cells, the nutrients that make up our bodies comes from life, the ecosystem services that clean our water and air are all provided by, Nature. The machines that we’ve built, the technology we thrive upon is nothing but Biomimicry

Biomimicry is an approach to innovation that seeks sustainable solutions to human challenges by emulating nature’s time-tested patterns and strategies.

We humans have been doing Biomimicry for a while. There are a ton of examples which are a result of this phenomena. For the sake of this article, we would limit it to 3 case studies namely (a) Aviation (b) Railways © Computer Science — Artificial Intelligence to illustrate the evolution of AI as a result of Biomimicry.

Aviation

The aviation industry is hugely indebted to nature. Airplanes fly by mimicking the aerodynamics of a bird. During the initial days of the aviation industry,

commuting on a plane was only possible for people who can afford. Because building flying machines is one thing, building efficient flying machines is another thing. The initial aircraft designs meant good enough to hit the skies. However, it wasn’t good enough for transatlantic flights or a commercial passenger plane. Many design challenges arose from minimizing drag, limited thrust, improving stability and balance of the overall aircraft structure. All of these challenges meant in-efficient fuel usage which equated to covering lesser distance. So the need of the hour was to have a structure which will alleviate some of the design flaws. From the research, it was clear one of the critical structural design change they have to improvise was wings. As part of the research different wing profiles were tested.

The diagram on the left shows a multitude of wing patterns.

Aviation researchers looked-up-to birds again to solve this problem. The curved wings of birds generated more lift than flat ones. Through the curvature airfoil design for the wings, airplanes can behave more similarly to birds. It improved their performance by less drag, which translated into fuel efficiency. Higher fuel efficiency also meant affordable planes which enabled the mass consumer market for aviation.

Advances in the avionics have led to improving that design even further. Supercomputers were used to study the bird’s anatomy and mimic the design on airplanes. Engineers have used a supercomputing technique that simulates natural selection to design the internal structure of an aircraft wing from scratch. The resulting blueprint is not only lighter than existing wings, but it also resembles natural formations, such as bird wing bones, that are not present in modern planes. The organic-looking product is as stiff as a conventional aircraft wing but lighter, which could save up to 200 tonnes of fuel per year per aircraft. Additionally, the design tweaks from fixed wings to more dynamic (flappable) in nature has added capabilities to perform complicated maneuvers. Which are well suited for the military aircraft that thrive upon to carry out aggressive, defensive operations.

The V-formation is something very widely seen in airforces across the globe. If you’ve been ever wondering why military aircraft fly in a V-formation, it is an efficient flying formation borrowed again from birds. Birds have been able to boost the distance they’re able to fly by more than 70 percent through the use of the V-shape. Scientists have discovered that when a flock takes on the familiar V-formation when one bird flaps its wings it creates a small updraft that lifts the bird behind. As each bird passes, they add their energy to the stroke helping all the birds maintain flight. By rotating their order through the stack, they spread out the exertion. There are reports that some military aircraft flew in a V-formation when they were low on fuel during the second world war.

Railways

In Japan, when the bullet trains got first introduced. The trains could notch up speeds of 200 mph (300 km/h), but the sound levels exceeded environmental standards. One source of noise was an atmospheric pressure wave forced in front of the train as it traveled through a narrow tunnel, creating a loud “tunnel boom” at the exit. It was called the piston effect. The sonic boom could be heard by residents at 400 meters away. This problem was particularly troublesome as it was due to both the geometry of the tunnel and the speed of the train.

An engineering team was tasked to design quieter, faster and more efficient structure of the train to keep noise levels under 70 dBa. Eiji Nakatsu was the head of the technical development of the engineering team. He was a bird watcher.

Inspired by Kingfishers unique beak that helps it to swoop down to catch prey. Owl’s stealthiness (noiseless flying pattern) and the Adelie penguin whose smooth body allows it to swim and slide effortlessly. Nakatsu redesigned the bullet train nose, pantagraph rig and pantagraphs supporting shaft respectively.

The redesign reduced the sonic boom effect and allowed the train to run at higher speeds and still adhere to the standard noise level of 70 dBa. The new Shinkansen 500 had 30 percent less air resistance than its predecessor. Energy consumption reduced proportionally.

A measured actual train run (maximum 270 km/hr) showed a 13 percent reduction in the power that had been needed by the predecessor 300 series. On March 22, 1997, JR-West put the 500-Series Shinkansen electric train into commercial service. The train was able to run at 300km/h at its maximum, a world speed record at the time, and met the stringent noise standard.

Computer Science — Artificial Intelligence

The evolution of artificial intelligence goes back as early as when the first computer was built. There is a difference of opinion on when was the computer invented and by whom. However, as the saying goes “Necessity is the mother of invention”. Computers got prominence during the world war 2 and post world war era. During the world war 2 British mathematician, “Alan Turing” built a machine that changed the course of the world war 2. Helped crack Nazi Germany’s ‘Enigma’ code and laid the groundwork for modern computing and artificial intelligence.

Artificial Intelligence is a branch of computer science that mimics biological(Human, animals or any other living organism) intelligence in machines.

Computers, to begin with, were intelligent. They did what they’ve been programmed to do. As technology evolved and machines (Car, Aircon, Refrigerator etc) got digitalized computational capability was a given thing in them. Additionally, the internet brought a whole new dynamics which enabled the machine to machine communication. Communication by itself isn’t meaningful when it’s not performed for a relevant context. In order to do that machines have to learn a bit more, not just about them, but also their surroundings (which have not been pre-programmed) to perform a contextually(present, future) aware action.

This is a common trait in humans and any other living organism. In order to perform such contextually aware actions. We heavily rely on new learnings and past experiences. This is a big piece of the equation that machines don’t have. The art of retro-fitting a biological behavior(learnings — new and from past experiences and acting upon them) in machines is what artificial intelligence is all about. So how do we do that?

Natural vision is mimicked by cameras. Hearing with the help of microphones and other natural sensory aspect is replaced by sensors which combined together generates a huge amount of data. As machines by themselves don’t have cognition, the huge set of data which have been generated doesn’t do anything useful. To make sense of the data and learn from it computer algorithms come in handy. Here in this article, we’d talk about one such prominent algorithm that illustrate biomimicry and enables artificial intelligence in machines (Cars, Computers, Robots etc) which is as follows:

Genetic Algorithm

Genetic Algorithms (GAs) as the name suggests is based on the concepts of natural selection and genetics. GAs are a subset of a much larger branch of computation known as Evolutionary Computation.

Genetics Algorithm start from a group of solutions (i.e initial population). Those solutions are combined to produce offsprings — the next generation of probably better solutions. New solutions are made from old ones using crossover and mutation just like nature does

The video below illustrates the working of Genetic Algorithm

I hope by now you might have a fair bit of understanding of GA. Now let’s look at some of the application of GA if not all of them.

Applications of Genetic Algorithm

• Automotive Design — design composite materials and aerodynamic shapes for race cars and regular means of transportation (including aviation) can return combinations of best materials and best engineering to provide faster, lighter, more fuel-efficient and safer vehicles for all the things we use vehicles for. Rather than spending years in laboratories working with polymers, wind tunnels and balsa wood shapes, the processes can be done much quicker and more efficiently by computer modeling using GA searches to return a range of options human designers can then put together
• Encryption and Code Breaking — On the security front, GAs can be used both to create encryption for sensitive data as well as to break those codes. Encrypting data, protecting copyrights and breaking competitors codes have been important in the computer world ever since there have been computers, so the competition is intense.
• Trip, Traffic and Shipment Routing — Your google maps or apple maps the shortest route is a great example of a genetic algorithm. Maps derive an optimal route based on the past routes. Which very well reflects the working of GA. Also, the most efficient use of transport for shipping, even to including pickup loads and deliveries along the way is another application of GA.
• Evolvable Hardware — Evolvable hardware applications are electronic circuits created by GA computer models that use stochastic (statistically random) operators to evolve new configurations from old ones. As the algorithm does its thing in the running model, eventually a circuit configuration will come along that does what the designer wants. Think of reconfigurable circuits in something like a space robot. It could use a built-in GA library and simulator to re-design itself after something like radiation exposure that messes up its normal configuration or encounters a novel situation in which it needs a function it doesn’t already have. Such GAs would enable self-adaptation and self-repair.

Final Thoughts

• Biomimicry is an approach to innovation that seeks sustainable solutions to human challenges by emulating nature’s time-tested patterns and strategies.
• Artificial Intelligence is a result of biomimicry
• Artificial Intelligence is a branch of computer science that mimics biological(Human, animals or any other living organism) intelligence in machines.
• Biomimicry is not only used in computers. However, it’s used across different domains to solve various problems.
• You could look at nature as being like a catalog of products, and all of those have benefited from a 3.8 billion year research and development period. And given that level of investment, it makes sense to use it

I hope this gave you some perspective about biomimicry and artificial intelligence. Stay tuned for more articles. Please feel free to leave your feedback.