Making sense of Embryomorphic Engineering and Artificial Embryogeny

You may understand the concept of artificial embryology by first grasping the concepts in the sequence of 1) evolutionary algorithms, artificial development, and finally artificial embryology, which itself is a form of embryomorphic engineering.

First you must understand what an evolutionary algorithm is:

The video above illustrates what an evolutionary algorithm is. To provide further illustration, certain algorithms such as the algorithm to allow a machine intelligence to do vision, requires an evolutionary process to continuously evolve sight. Sight cannot simply be a concrete thing, but similar to how a baby must use their eyes to develop a sense of sight, it is similar for a machine which must use it's vision to develop it's ability to see accurately. Over time the accuracy and detail increases due to the evolutionary process.

Short video:

Long video:

A technique of automatic algorithm invention allows for the machine intelligence to apply a genetic process or evolutionary process, to evolve sight. It works where for each try which could be seen as a candidate solution, there is a process which measures fitness of each candidate, and over time the best algorithm or best try is evolved as the most fit. There needs not be a programmer to write new algorithms to refine the vision for the computer because the computer uses a genetic algorithm to refine and evolve a best fit algorithm for sight. In essence, the more images the computer sees, the more accurate the visual algorithm is trained to get.

That is a very dramatically simplified rough explanation, but the videos of course provide some more details. The point is that the software mimics how the brain, or how genetics soft problems. It's very much like a trial and error process, it's a machine learning process.

The hardware side of this is neuromorphic computing (brain based chips) which can better take advantage of this kind of evolutionary software.

Artificial development

Artificial development is a nature inspired evolutionary computation based development method. Specificially, it is a design based on biomimicry for the purpose of engineering, such as software engineering. Rene Doursat is a trailblazer in the field of artificial development and his current breakthrough is called morphogenetic engineering. Artificial Embryogeny is bio-inspired, where by studying how an embryo works, the processes can then be used to improve development processes. An artificial embryo is very similar in approach to an artificial neuro-network approach, it's inspired by a biological process.

Morphogenetic Engineering: René Doursat

Alan Turing's final paper was on the topic of morphoegenesis. Morphogenesis explains the process of why a leopard has it's spots, which allows for advances in the understanding of camouflage. Alan Turing was studying the chemical basis of these patterns in his final paper. Morphogenetic engineering is about understanding and engineering self organizing programmable complex systems. A complex system is a generic term, but by understanding complex systems you can understand better complex adaptive systems, and a brain could be seen in this way, or a blockchain can be seen in this way.

EvoDevo

Evolutionary development and morphogenetic robotics

New industries for robotics are being developed. For anyone who has heard of cloud robotics, where the brian of the robot exists in the cloud, morphogenetic robotics are bio inspired and can take advantage of this cloud robotics architecture.

Cloud robots are basically internet connected robots. However, bio inspired methodologies can have a dramatic influence on robotics and are worth the time for further investigation in my opinion for anyone interested in the subjects of evolutionary development or robotics.

References

Chavoya, A. (2009). Artificial development. In Foundations of Computational, Intelligence Volume 1 (pp. 185-215). Springer Berlin Heidelberg.

Doursat, R., Sánchez, C., Dordea, R., Fourquet, D., & Kowaliw, T. Embryomorphic Engineering: Emergent Innovation Through Evolutionary Development.

Doursat, R., Sayama, H., & Michel, O. (2013). A review of morphogenetic engineering. Nat Comput, 12, 517-535.

Turing, A. M. (1952). The chemical basis of morphogenesis. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 237(641), 37-72.