Salamander robot walks, swims and sheds light on evolutionary step from sea to land

A team of scientists have developed a robot salamander that walks and swims using an electronic spinal cord. The robot provides us with clues about how our ancestral animals made the evolutionary leap from the sea to the land.

Moving robots are becoming more and more advanced, from Honda’s astronaut-like Asimo to the dancing Robo Sapien, a perennial favourite of Christmas stockings.

But these advances are still fairly superficial. Most robots still move using pre-defined programmes and making a single robot switch between very different movements, such as walking or swimming, is very difficult. Each movement type would require significant programming effort.

Salamanders inspired a robot that tells us about the transition from sea to land.But robotics engineers are now looking to nature for inspiration. Animals of course, are capable of a multitude of different styles of movement. They have been smoothly switching from swimming to walking for hundreds of millions of years, when our distant ancestors first invaded the land from the sea.

This ancient pioneer probably looked a fair bit like the salamanders of today’s rivers and ponds. On the land, modern salamanders walk by stepping forward with diagonally opposite pairs of legs, while its body sways about its hips and shoulders. In the water, they use a different tactic. Their limbs fold back and they swim by rapidly sending S-like waves down their bodies.

Both of these movements, as in all back-boned animals, are controlled by bundles of neurons called central pattern generators (CPGs). These bundles run down either side of the animal’s spine (its body CPG) and in each of its four limbs (its limb CPGs).

The CPGs produce rhythmic movements in muscles, by sending them carefully timed pulses of electrical signals. The brain is a casual bystander in this process, stepping in only to tell the CPGs to switch from a walking to a swimming rhythm, or vice versa.

By directly stimulating the brains of salamanders, Jean-Marie Calguen from the University of Bordeaux, managed to trigger the switch between walking and swimming gaits. With low levels of stimulation, the hapless animal made walking movements, and at higher levels, it tried to swim.

Salamandra Robotica walks its way towards Lake Geneva.Calguen, along with Auke Ijspeert from the Ecole Polytechnique Federale de Lausanne, came up with a model for how this switch works and tested it by building a robot salamander. The metre-long and grandiosely named Salamandra Robotica was designed to mimic its biological counterpart.

Its movements are controlled by an ‘spinal cord’ that uses electronic CPGs to control its body and limbs. Like a real salamander, these are overseen by signals from the robot’s ‘brain’ – in this case, a wireless human-controlled laptop.

Ijspeert and Calguen used different CPG programs to control the robot’s body and limbs. When the robot receives any stimulation from its laptop brain, its body CPG produces the S-like body waves used by swimming salamanders.

At low levels of stimulation, the limb CPGs overpower the bodily ones, and the robot walks. But the limb CPGs cannot cope with higher levels of stimulation and switch off, leaving the bodily CPG free to start a swimming motion.

This model was a success. When they tested Salamandra Robotica on the shores of Lake Geneva, Ijspeert and Calguen found that their robot reproduced the same swimming and walking gaits seen in living salamanders, abruptly switching between the two depending on how much stimulation its CPGs received.

Acanthostega, an early invader of land, probably walked like modern salamanders.The model shows one way in which evolution could have modified an aquatic animal’s movements to a walking way of life. This was a key evolutionary step and provided the impetus for the spread of life from sea to land.

Salamandra’s success also shows that the studies of robotics and biology can successfully work together. Robots can be used to test biological ideas, while biology in turn can inspire successful solutions to engineering problems.

The robot salamander is part of a new era in robotics, where robot movements are controlled by artificial nervous systems.

 

Reference: Ijspeert, Crespi, Ryczko & Cabelguen. 2007. From swimming to walking with a salamander robot driven by a spinal cord model. Science 315: 1416-1419.

(Salamander photo by Marek Szczepanek. Photo of Salamandra Robotica from EPFL. Drawing of Acanthostega by Arthur Weasley.)

Technorati Tags: , , , , , , ,

New learning robot adapts to injuries

The field applications of modern robots are limited because they are largely unable to cope with damage and injuries. But a new robot called Starfish could change all that – it is programmed to learn about itself and adapt to new situations, with no pre-built contingency plans.

I am walking strangely. About a week ago, I pulled something to my left ankle, which now hurts during the part of each step just before the foot leaves the ground. As a result, my other muscles are compensating for this to minimise the pain and my gait has shifted to something subtly different from the norm.

In similar ways, all animal brains can cope with injuries by computing new, often qualitatively different, movements to compensate. Because this isn’t a conscious process, we often take it for granted. But by looking at how difficult it is to program a robot to do the same thing, we can get a sense of how hard it actually is.

The Starfish robot explores its own body. Robots have been used for years to perform structured, repetitive tasks. As engineering has advanced, their movements have become more and more stable and life-like.

But they still have severe limitations, not the least of which is inflexibility in the face of injury or changes to its body shape. To put it simply, if a robot’s leg falls off, it becomes as useful as so much scrap metal.

So for robots, adaptiveness is a desirable virtue. Modern bots can independently develop complex behaviours without any previous programming but this requires trial and error and lots of time.

Josh Bongard and colleagues at Cornell University , New York, have solved this problem by developing a robot (see right; image from Science magazine) that continuously assesses its body structure and develops new ways of moving if anything changes.

It differs from other models in that it has no built in redundancy plans, no strategies for dealing with anticipated problems. It’s simply programmed to examine itself and adapt accordingly.

The concept of a robot that can adapt to new situations is often the precursor to nightmare scenarios in many a science-fiction film. So it is fortunate that Bongard’s robot isn’t armed or threatening, but instead looks more like a four-armed starfish.

Each arm has two joints, and sensors that record the angle of these joints, and the tilt of the arms.

At first, Starfish performs some experiments. Humans have an instinctive understanding of how our body parts connect with each other, but this sense, called kinaesthesia, must be programmed into robots. Bongard’s robot doesn’t need that – it can work out its structure on its own.

It does this by performing random actions and using an array of sensors to see what these do to its body. It then creates several ‘self-models’ – representations of how its body is joined together – in the same way that a forensic scientist pieces together how a crime occurred based on the evidence.

Starfish then compares the models and performs actions designed to distinguish between them. After several rounds of this, the robot has a fairly accurate idea of how it’s built, what sorts of things it can do, and which parts it needs to move to do them. If it’s given an instruction, such as ‘move forward’, it can plan the best way of doing that.

However, the robot detects something funny that goes against its self-model, it initiates the whole process again. If it’s leg falls off, it notices, re-creates its picture of itself, and plans new behaviours to cope with its new situation.

These abilities will be instrumental in the future of robotics. Robots will become exponentially more useful if they can respond to new environments, or cope with the bodily changes that happen when they grasp a tool or suffer damage.

They can be deployed to unstable disaster sites to help with recovery, or to the depths of space for exploration. They may even give us an insight into how the human brain develops self-awareness and adapts to new situations.

Bongard, Zykov & Lipson. 2006. Science 314: 1118-1121.
Technorati Tags: , , , ,