Boko Park-Popular Science: A paper recently published in Scientific Robotics describes this valve. In addition to Rothemund and Preston, this research was conducted by Alar Ainla, Lee Belding and Sarah Kurihara (from the Department of Chemistry and Chemical Biology), Zhigang Suo (from Kavli Bionano Institute of Science and Technology) and Whitesides***. People have made many different types of soft robots ... Finally, all these robots are controlled by hard valves. Our idea is to implant these control functions into the robot itself, so that we no longer need these hard external parts. This valve combines two simple ideas: first, the membrane is similar to a "popper" toy, and second, when you twist these pipes, it is like you twist the garden hose to stop the water flow. The valve demonstrated by Preston and Rothemund is built in a cylinder, which is separated by a silicon film to form an upper cavity and a lower cavity.
When it falls on an object, the valve closes and the gripper starts automatically. Photo: Harvard University
Pressurizing the lower chamber forces the membrane to eject, releasing the pressure and ejecting it back to the "static" state. Each chamber also includes a tube that can be twisted to effectively open or close the valve when the membrane changes direction. No matter which direction you face, the upper tube or the lower tube will be knotted, so when you eject, the bottom tube will bend and no air will pass through the bottom tube. When the film is ejected, the upper tube is twisted, the lower tube is disconnected, and air can flow through the lower tube. We can switch back and forth between these two States ... switch the output. Preston and thurmond said that in some ways, this valve represents a new method of soft robot. So far, most of the work in this field has focused on functional manufacturing robots, which can grab or act as soft surgical tractors. Luo thurmond and Preston believe that this valve is a key component and can be used in any number of devices.
This method is suitable for any soft actuator, and it doesn't answer the question of how to make a fixture, but to say the least, many soft robots have the same principle of inflation and deflation, so these robots can use this valve. Preston and thurmond can automatically adjust the valve to perform some actions, such as catching an object. In one demonstration, the valve was designed as a multi-finger fixture, but a small vent was added to let the air pressure release from the chamber at the bottom of the valve. However, when the clamp is lowered onto the tennis ball, the vent hole is closed, which leads to pressurization of the bottom chamber, activation of the valve and operation of the clamp. So this technology integrates this function into the robot. Someone made a clip before, but someone always stood there and watched the clip get closer. This is automatic.
The team was also able to establish a "feedback" system, which, when providing a single stable pressure, would cause the valve to oscillate rapidly between different states. Essentially, the system transmits air pressure from the upper chamber to the lower chamber. When the valve pops up to the raised position, it cuts off the pressure, allows the bottom chamber to vent, releases the pressure, and returns the membrane to the lowered position to start the cycle again. We take advantage of the fact that the pressure that causes the membrane to turn up is different from the pressure required for the membrane to turn down, so when we feed back the output to the valve itself, we will get this oscillation behavior. Using this behavior, the team can build a simple inchworm robot that can move based on a single valve that accepts a single input pressure. Under a constant pressure, we can make such a walking action, but we can't control this walking at all-we just need to input a pressure and it can walk on its own.
More work needs to be done in the future to further improve this valve so that it can adapt to different uses and different geometries. This is just a demonstration made of film. There are many different geometries that show this bistable behavior ... so now we can consider designing this robot and making it suitable for robots, depending on the application you want. Preston also wants to explore whether this valve, which is always in one of two states, can be used as a transistor to form a logic circuit. To some extent, it is a bit like a transistor. You can input pressure and then convert the output ... in this sense, we can regard it as a part of a completely soft computer.