Today is Charles Darwin's 200th birthday and this month also marks the 150th anniversary of his book, On the Origin of Species. As might be expected there are shindigs and events all over the world. There's even a Darwin Look-Alike contest at ASU and a Darwin Google logo for the occasion. Other organizations and website are focusing on how other fields of science were changed by Darwin's theory of evolution. But not us, no siree. We're going to remind you how much evolution has impacted robotics. While the direct application of evolutionary algorithms is obvious, there are other less obvious connections. For example, scientists are about to release a rough draft of the Neanderthal genome, pieced together from 38,000 year old bones found in Croatia. By comparing the Neanderthal genes to those of chimpanzees and humans, we'll learn a great deal about how our brains evolved. This in turn will suggest new approaches for creating intelligent robots.

Nearly every field of human endeavor is affected by or even depends on evolution in some way. Because of this, Darwin's explanation of evolution proved enormously useful to science in much the same way as Newton's theory of gravity. In both cases a fact was explained by a new, approximately correct theory. Just as Newton's theory was refined by Einstein into the general theory of relativity, Darwin's theory of evolution has been refined into what's today know as modern evolutionary synthesis. But in both cases, later refinements in no way diminsh the importance of the original work.

As with many previous scientific advances, both the fact and theory of evolution were initially considered controversial because of a perceived threat to religious beliefs. Today, most mainstream religions have accepted and even embraced evolution. The Anglican Church has added a special section to their website for Darwin's birthday including the essay, Good Religion needs Good Science, by Rev Dr Malcolm Brown. The Presbyterians are having "Evolution Sunday" on February 15. The Episcopal Church has affirmed its acceptance of evolution since 1982. The Clergy Letter Project is having their annual Evolution Weekend. Rev. Gary McCaslin of First Baptist Church, Painted Post, NY offers his evolution sermon, An Informed Faith ~ the Union of Good Religion and Good Science (DOC format), as an example for other churches.

Even the Vatican, never known as an early-adopter when it comes to science, recently pronounced evolution compatible with the Christian faith and are holding their own conference called Biological Evolution: Facts and Theories. Francisco Ayala, an evolutionary biologist and former Dominican priest, will give a lecture in Dallas, TX at the Southern Methodist University titled Darwin's Gift to Science and Religion on Feb 20th. Ayala will also be speaking next month at the Vatican conference. The Pew Forum has an interesting page on the stance of various religions and sects with regard to evolution.

It's not all good news, of course. There are still highly vocal holdouts such fundamentalist Islamic and Christian groups who reject evolution as theory and even deny evolution as fact. Often these are the same groups hanging on to other historic concepts such as belief in geocentrism, Cartesian dualism, and even a flat Earth. An Australian news article, the Right should warm to Darwin, points out the irony that many of these anti-evolution groups, particularly those in the western world, simultaneously embrace seemingly Darwinian ideas such as order emerging spontaneously from chaotic markets or free societies. In a NY Times interview Francisco Ayala had this to say about how he deals with students who pose religious objections to evolution in his basic biology class at the University of California:

With Catholics, I take out the Pope's address to the Pontifical Academy of Sciences in October 1996 where he endorses evolutionary teachings. If the students are Christian fundamentalists, I tell them that there are many Protestant theologians who agree with evolution. I say that evolution, in my view, is not only NOT anti-Christian, but the idea of special design, which many fundamentalists adhere to, might be -- because it teaches the view of God that is blasphemous. The Special-Design-God is a God who messes up. Think about all the backaches, infected wisdom teeth and painful childbirth that exist because we humans evolved incompletely! ''Do you think God is absent-minded?'' I ask them.

So, OK, granted evolution isn't as controversial as it once was, what's any of this have to do with robots? The importance of evolution to many fields of science is immediately evident by taking a look at the recent flood articles and lectures on the subjects. For example, the College of William & Mary is offering a series of lectures today discussing the importance of evolution in the following fields: medicine, public health, developmental biology, anthropology, geology, psychology, organizational ecology, linguistics, even theology, morals, and ethics. That's hardly an exhaustive list but you get the idea.

It's easy to see how many of those might apply directly to artificial intelligence and robotics. Anyone studying computer science is aware that modern software algorithms owe a lot to evolution whether indirectly, such as the inheritance and polymorphism of object oriented programming, or directly as in evolutionary algorithms. These evolutionary algorithms implement the real thing, with varying amounts of similarity to the form existing in nature. Far from being an obscure branch of computer science, you probably rely on evolutionary algorithms every day, perhaps without even knowing it.

Genetic algorithms are trading stocks in your mutual funds, making your video games more realistic, and designing circuits for use in consumer gadgets. Do you use a cellphone? Several cell service providers have used evolutionary algorithms to optimize their network design or determine call admission policies. Ever flown on a plane? Genetic algorithms help manage air traffic control systems. A recent techradar article notes that Volvo relies on other evolutionary algorithms to schedule resources while a company in Scotland relies on evolutionary software to manage the supply chain for 7 million barrels of whiskey.

Now, as you sit in your first class seat, drinking whiskey and checking mutual fund results on your smart phone, you're probably wondering, so what? Let's hear something specific that evolution has done for robots. Fair enough. Let's take a look at how evolution has directly impacted robotics. Here's a small sampling from our own story archives.

• researchers at Chalmers University in Gothenburg, Sweden have used evolution to develop bipedal walking mechanisms for robots. Designs were initially evolved in a simulated environment and then tested on robot hardware in the real world. The researchers found that evolution could provide stable gaits that outperformed hand-tuning of human-created software.
• French researchers have applied evolutionary optimisation to stereo vision analysis, resulting in a low-cost solution to the problem of robot obstacle avoidance.
• In an ironic reversal, here's one case of robotics technology illuminating the results of natural cellular evolution. Researchers at Princeton studying cellular proteins discovered a set of proteins that implement a bang-bang controller of the type commonly used in conventional robotics for navigation. In the cell, the controller set a deterministic course for evolution at the cellular level, guiding random mutations in a way that correct damage and imbalances.
• Students at NMU simulated robots that evolve reproductive behaviors selecting for capabilities related to communication, learning, and abstraction. Their goal was to show that autotrophic reproduction would potentially result in intelligent robots.
• Researchers at the Hong Kong Baptist University used Multiphase Genetic Programming as a learning mechanism which allowed a real Sumo robot to evolve competitive maneuvers.
• Researchers at Chalmers University used evolutionary programming algorithms to develop flapping techniques in a winged robot, demonstrating that viable flapping motions can evolve naturally.
• Adrian Thompson implemented genetic algorithms directly in hardware using self-modifying FPGAs to evolve circuits that defy human design rules. In some cases researchers still don't understand how the resulting circuits work.
• Julian Togelius demonstrates a combination of evolutionary computing, neural networks, and subsumption architecture that allows simulated robots to learn simple behaviours such as conditional phototaxis and obstacle avoidance.
• CMU researcher M. C. Martin describes the use of genetic programming to evolve a robot vision system for reactive obstacle avoidance.

If evolution didn't work, none of these things would be possible. More importantly, if we didn't understand how evolution worked, none of these things would be possible. It was Charles Darwin who started us down that path of understanding. So Happy Birthday and enjoy the cake!