Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Shape changing robots

  1. Jul 6, 2006 #1
    I once had a chat with a Prof. at school, who talked about amoeba like robots. Since I am not so much into robotics, but nevertheless found the concept very amusing, I'd like to get some feedback on what are the feasibilities of the ideas below (all of which are just random thoughts supported by some half-baked knowledge of engineering I have):

    Let's start with a technical definition: Shape changing here can be best defined as ability to vary surface area (assuming the volume remains constant). Thus SAmax/SAmin is the more for an amoeba than say, a snake.

    1. Controlled polymerization: The movement in an amoeba is largely due to chemical process. A robot design directly flowing from this concept can use chemicals to achieve motion/change in shape. For e.g. a plastic membrane with semi-solid form of the same material inside directionally heated and cooled to achieve a change in shape. The heating system could be located outside or inside the robot.
    2. Non-modular structure: This is contrary to the quantized approach of making several robots that can reconfigure. (At the same time, there can always be a combination of such robots to create a reconfigurable robot). Imagine a robot made of telescopic tubes attached around a sphere. By controlling the length of these tubes the shape of the robot can be changed. At a very basic level, this can be like three Stewart platforms attached together with their base such that the outward shape can be altered. (See fig.)
    3. Controlled fluid flow: By controlling the amount of fluid (let's say, air) in small bags, a robot can achieve some degree of shape-changing. This would essentially involve sucking up air from surroundings and selectively filling up small bags to achieve a particular shape.
    4. Reconfigurability(building on the quantized approach): There is a lot of research going on reconfigurable robots. Usually this involves using several similar robots to achieve a reconfigurable design. The limitation and thus shape-changing characteristic for these robots is governed by the size of actuators, sensors, and coupling mechanisms. Is it possible to radically change the interfaces i.e. moveable faces, magnetic couplings, etc.

    Thanks for tolerating up till here :smile: Now, what do you think?

    Attached Files:

  2. jcsd
  3. Jul 6, 2006 #2


    User Avatar
    Gold Member

    They all sound reasonable to me, although some of the examples might be of limited usefulness.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook