Modularity in Robotics

modularity-in-robotics

The modular design approach, or modularity, subdivides a system into smaller and similar parts called modules, which can be customized, upgraded, repaired, scaled, and reused. The concept of modularity is not new and can be traced back to World War II, when Otto Merker investigated how modularity could be implemented in constructing German U-boats. IBM launched modular architecture for personal computers in 1980, and today, modularity is a well-known term in the world of robotics.

In the late 1980s, the concept of applying a common connection mechanism to the entire robot was introduced by Toshio Fukuda with the CEBOT (short for cellular robot). The early 1990s saw the initial developments of modular robots in the form of lattice reconfiguration and chain-based systems, followed by hybrids. The lattice configuration can build two-dimensional (2D) and three-dimensional (3D) structures by arranging the modules to form a grid, just like atoms forming complex 3D molecules or solids, while chain-based configuration can create a chain of modules; for example, a four-legged robot can be thought as five chains. A chain acts as the main body and another four chains conform the legs.

Moving Modularity beyond the Research Labs

Modularity in robotics is highly research oriented, but attempts are underway to move the approach beyond research labs to the commercial sector. The research teams from the MoCoTi EU project and Myorobotics have developed a robot prototype that learns how to actuate its own limbs, which can be easily duplicated. The robot consists of an artificial brain that controls a tendon-driven robotic arm, which is perhaps a step toward building low-cost humanoids. According to Christoph Richter, one of the leading researchers for MoCoTi from the Technical University of Munich, “This is possible because its modular design permits a relatively efficient mass production.” The next step is to design an artificial cerebellum to control the orders from the locomotor system. The researchers selected a neuromorphic computer platform called SpiNNaker developed at the University of Manchester. A single chip module of a neuromorphic computer platform can manage a network of 10,000 neurons in real time and is far superior to chips in desktop computers. Thousands of such modular chips can be interconnected to simulate sizeable, brain-scale neuronal networks.

A very well known and often quoted example of a commercially available modular system is Lego. Its Mindstorms series of kits contain software and hardware to create customizable, programmable robots. They include an intelligent computer control module called a Brick that controls the system, a set of modular sensors and motors, and Lego parts from the Technic line to create the mechanical systems. The third-generation kit in Lego’s Mindstorms line, Lego Mindstorms EV3, was released on September 1, 2013. Some robot competitions using this set include the First Lego League and the World Robot Olympiad.

Both Startups and Big Corporations Are Showing Interest

Several startups in robotics are positioning themselves to offer products centered around the idea of modularity; examples include Cubelets and littleBits. Cubelets consist of cubes with various abilities ranging from sensing proximity to physical movement, and littleBits is a platform of easy-to-use electronic building blocks. Both startups and big corporations are showing serious interest. SCHUNK, one of the world’s largest manufacturing companies focused on gripping and clamping technology, offers modular systems to build robots such as Care-O-bot. Recently, at the Tokyo Motor Show, Honda demonstrated its RoboCas concept, a small-sized modular electric mobility robot. Honda envisions the body of the robot as being customizable and modular, allowing for different types of carrying spaces, such as a portable café.

Modularity’s Future in the Commercial Realm Faces Addressable Challenges

Modularity combines the benefits of standardization with customization, potentially offering tremendous economic and functional advantages. While there has been growth in the amount of research and experimentation of modular robots in space and deep-sea exploration, construction, search and rescue missions, disaster relief, and other uses, it is still difficult to identify specific commercial applications where modular design benefits can be demonstrated in both the short and long term. Even though the robustness and performance of modular robotic systems have been continuously improving over the past few years, there are doubts about optimizing such systems for a high degree of performance for demanding applications. When the challenges involving optimization, adaptability, and control algorithms are solved, the future prospects of modularity in robotics will certainly look much brighter than current expectations.

Comments are closed.