Living systems and the way they perform their functions, has been the reference model for building many technical systems, giving rise to bio-inspired technologies. In the computing technologies arena, while it has been more natural to build domain specific bio inspired algorithms and software systems, there has also been breakthroughs in the development of bio inspired computing hardware as well.
I have been obsessed with computing for a long time, since I learned about zero and one in school and that we could translate everything by these two digits. That fascination was the reason I did my first post graduate degrees in electrical and computer engineering focusing on computer hardware. Hardware is of course in accompany of software that makes the computing environment. The way we develop software algorithms, at first step is the simple automation of the way we would process the data and do operations manually. This automation certainly provides speed of operation and processing. However, it is only when we design our software algorithms the way nature performs its function, we can reach a level of exponential speed and optimisation based of the perfection of the nature.
Bio Inspired Algorithms
Many consumer applications and industrial use cases are generating massive amount of data. Processing huge amount of data could consume huge amount of processing, if the algorithms and software are not optimised for the task. Given the complexity of these applications and level of uncertainties they deal with from dynamic problem definitions, various changing constraints, lack of complete information as well as limited computation capacity specially in devices, they requires different way of processing that could cope with this environment.
Normally there are two aspects to developing this generation of algorithms, one is about creating and continuously improving a correct bio inspired algorithm, the other is about experimenting with application of these algorithms and evaluating the results. There is also the question of investment and the reusability of these algorithms that are normally created for one specific domain. Some famous example of bio inspired algorithms are neural networks, genetic algorithms, particle swarm, ant colony optimisation and flower pollination. While neural networks are more widely used, other algorithms are not yet so. Cognitive computing on the other hand uses the same way human thought process handles a complex situation to provide solution for a question for which the answer maybe ambiguous and uncertain.
Bio Inspired Computing Hardware
The biology has also inspired build up of the computing hardware, from the way we build general purpose computers to more domain specific ones. For example DNA computing, neuromorphic computing and quantum computing. Neuromorphic computing is based on neural networks with the same goal to make the computer smarter. DNA computing is based on DNA principles aiming for speed while quantum computing also aim at faster computing but uses the quantum principles.
Different Computing for Different Objectives
While the bio inspired algorithms and computing hardware are showing promising result, it is also evident that not only there is cost associated with such computing, they also would show their true value only for specific set of problems.
While ant colony and genetic algorithms are best for management and operational situation, Leaping frog algorithm is suitable for rediscovery, and artificial plant development is suitable for theory development.
In the computing hardware categories, neuromorphic computing is best suited for design of artificial neural systems in intelligent machines, such as vision systems, auditory processors, and autonomous robots, which are behaving based on those of biological nervous systems. DNA computing by having parallel computing in its nature provides speed and therefore suited for optimisation problems among other things. Among the specific problems suited to be solved by quantum computers are search, cryptography and linear equation.
While these are only few example, the point is that different computing paradigms are good for different objectives and not necessarily for the day to day usage of computing we have today.
Being inspired by biology to develop software algorithms and computing hardware, is the path we have always followed from the beginning of developing computing environment. In the normal computer the usage of storage, memory and processing unit is also inspired by the way the brain works, with long term memory, short term memory and processing of data. However, the advancement of understanding of biology and investing in multidisciplinary research can inspire the development of much more advanced technology mimicking the perfection of the nature itself. Looking forward I predict and hope for much more investment in multidisciplinary research combing optimisation and learning from one domain to another and specifically from biology into technology.