Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
SAF Archives  search ask the scientists in the classroom cool science
scientists from previous shows
cool careers in science
ask the scientists

Photo of Roger D. Quinn Roger D. Quinn as seen on Natural Born Robots: Roboroach

Click on Roger's photo to read a brief bio.

q I am a 15-year-old programmer who is very interested in robotics and trying to build my own robot. Can you suggest how I can get started? Thanks! (Question sent by Chris and many other viewers)

A There are Lego kits that are advanced enough to be very interesting. Randy Beer and Hillel Chiel (both at CWRU) started a course here that uses these kits. The really fun part for a programmer is that you can build a robot fairly quickly and then start working on software. Their students have implemented some complex behaviors in their Lego robots.

q How soon do you think you'll have a robot that can move as efficiently as a roach? And do you plan to add roach-like wings to your robot one day? (Question sent by Alicia)

A Truthfully, maybe never or at least a very long time. If we can make a robot that is half as agile as a roach, we will have a very impressive mobile vehicle. Because legged locomotion is a very difficult problem, we approach it by working on realizable goals. For this robot we are working on walking and climbing. For another cockroach robot now under construction, we are working on efficient running.

Wings for flying would certainly be good for locomotion over long distances. In fact, others are working on small flying vehicles (micro air vehicles or MAVS) based on dragonflies for example. We will monitor their work so that we might be able to incorporate into a future, roach-sized robot. We will continue to concentrate on legged locomotion. For example, one of our other robots will walk and jump based upon cricket.

Roy (the lead biologist on this project) points out that large cockroaches like the ones we studied for this project don't fly very well at all. Smaller ones do, but this might actually indicate a physical limitation for using insect-like flight for a heavy object like a roach or a 30 pound robot. Nevertheless, wings for flying would certainly be good for locomotion over long distances.

q I understand how you can make the robot move like a real cockroach, but won't the robot be more vulnerable than the insect to outside forces such as rough terrain because of the material it is made with? Because the metal is inflexible, won't it break easily? (Question sent by Cindy, Homestead High School, Cupertino, CA)

A Aluminum is a tough material. It is not brittle and does flex. However, you have a point about the animal's exoskeleton being well designed for its use. It is a complex, composite material with properties that vary throughout the structure. It seems to be stiff where it needs to be stiff and softer where it should be soft. Furthermore, its thickness varies for similar reasons. In comparison, our robot has a homogeneous material and therefore is not optimized for strength vs. weight, which is important for performance. The cockroach's designer is way ahead of us. We could concentrate on materials or actuators, but we choose to concentrate on legged locomotion.

q Please direct me to Internet sites that explain how scientists are using facts about insect locomotion to build robots. Thank you. (Question sent by KC and other viewers)

A Our site is We have links to other sites there. We should probably add some more in the near future. (e.g., Univ. of Portsmouth, Holk Cruse in Germany, Hirose in Japan, Shimoyama in Japan., and the Univ. of Illinois)

q We are doing a project on how people use geometry in different professions, and I am doing robotic engineering. I was wondering if you could tell me any aspects of your work that you think you use geometry to help you. (Question sent by Andrew, Brewster Academy)

A Geometry is the basis for mechanical design and is the foundation for computer aided design (CAD). We use AutoCAD (commercial software) to design our robots. Without this, it would take many years to design a robot like RIII. Also, in Roy Ritzmann's lab, they use high speed video to study the motion of the cockroachs' legs. They need at least two views to see the 3 dimensional motion of those legs. The math needed to resolve these two 2-D views into 3-D motion is also geometry.

q Can you please explain how the Roboroach is powered? (Question sent by several viewers)

A This robot has pneumatic actuators (air cylinders). High pressure air can enter at either end of the cylinder. When high pressure air enters one side of the piston, it drives the piston toward the other end of the cylinder and causes a joint to rotate. Valves on the "abdomen" of the robot distribute air to the cylinders driving the 24 joints. High pressure air is delivered to the valves by a hose connected to a compressor that is not on board the robot.

q Will you complete the roboroach with material to cover the skeleton? Will the material be similar to the real roach, or will you use different materials at different areas of the body (for example, harder material for the legs to grip the ground)? (Question sent by Aram M.)

A Your question points out one of the non-roach-like characteristics of Robot III. It does not have an exoskeleton that covers its actuators. It is built more like a vertebrate with a skeleton. We are now designing Robot IV. This one will have an exostructure and its actuators will be placed inside. Please see for a picture of a prototype leg. Because of this, Robot IV's mass distribution will be more similar to that of the cockroach and this should help it to run more like a roach. Yes, the cockroach has a complex material makeup (see another of our answers on this topic), but most or our robots' materials will remain aluminum for simplicity. We do use a soft, compliant material on the foot to reduce the shock of impact and to improve traction. Cockroaches have a compliant tarsus (foot).

q I know that the cockroach robot needs to remain lightweight, but it needs more power, so are there any new biologically inspired methods of making a robot move? I've heard about the development of synthetic muscle tissue, but is that realistic, or practical? (Question sent by Masumi W.)

A Legged robots are like aircraft: a higher power to weight ratio for their engines and a higher strength to weight ratio for their materials provide better performance. We use pneumatic actuators now instead of electric motors because of this. There is much ongoing research into artificial muscles and we would love to have them commercially available. We are moving from air cylinders to McKibben artificial muscles (also pneumatic). These were developed in the 1950's, but the materials for them are now readily available. If you do a search on the web you will find research groups working on artificial muscle (e.g. MIT, S.R.I.) They will become practical eventually and that will make our job much easier.


Scientific American Frontiers
Fall 1990 to Spring 2000
Sponsored by GTE Corporation,
now a part of Verizon Communications Inc.