David Brandt

A unified simulator for Self-Reconfigurable Robots

Generic simulation platforms such as player/stage are an essential tool in mobile robotics, but until now no similar platforms have been available for the field of self-reconfigurable robots. We here present a generic simulation platform for modular, self-reconfigurable robots: the unified simulator for self-reconfigurable robots (USSR). USSR is based on a physics engine, allowing simulation of both self-reconfiguration and dynamic interaction with the environment. The simulator is implemented as a framework that provides numerous components that can be combined to form new or existing modular robots, allowing easy experimentation: USSR currently includes support for the ATRON, Odin, and M-TRAN modular robots.

ATRON Robots: Versatility from Self-Reconfigurable Modules

Traditional fixed morphology robots are limited to purely functional adaptation and thereby to a limited range of applications. In contrast modular self-reconfigurable robots can dynamically and autonomously change both their function and morphology to meet new demands of changing tasks. Therefore, self-reconfigurable robots have the potential to become highly versatile. This paper documents and discusses the application versatility of self-reconfigurable robots in general and of the ATRON system in particular. We present a range of different self-reconfigurable robots assembled from the ATRON base module. The robots ability includes locomotion (snake, car, and walker), manipulation of objects (serial manipulator, conveyer belt) and autonomous change of functionality and shape (locomotion configurations, many-module shape-change). We also demonstrate the structure of simple anatomical building blocks (bones, muscles, etc.) which we envision can be assembled into more complex robots of future miniaturized modules.

A new meta-module for controlling large sheets of ATRON modules

In this paper we present a 2D meta-module for the ATRON robot, which simplifies the motion constraints significantly. The motion capabilities of the new meta-module is similar to that of previous sliding cube style modules with the addition of one extra action, which is shown to improve the motion capabilities of the modules greatly. In general our work shows that if three simple actions, which is implemented by our proposed meta-module, is implemented by a self-reconfigurable robot module or meta-module then the control of the robot will be simple. The improved motion capabilities allow us to use offline planning for large groups consisting of hundreds of meta- modules. The simple motion constraints of the meta-module further allows us to implement a distributed cluster flow locomotion gait, with the interesting property that the group of meta-modules self-organizes into a shape that makes the locomotion efficient.

Comparison of A and RRT-Connect Motion Planning Techniques for Self-Reconfiguration Planning

This paper presents a comparison between two different algorithms for self-reconfiguration planning on modular self-reconfigurable robots. The two algorithms are A* and RRT-Connect. A* has often been used for self-reconfiguration planning on different systems, but so far no work has been done to show that A* is an optimal or even a good choice for this task. The RRT-connect algorithm is adopted from traditional robot motion planning and is easily adapted to self-reconfiguration planning. The results show that RRT-connect is significantly faster than A* for difficult self-reconfiguration problems. In most cases RRT-connect generates longer paths that A* it is, however, possible to optimize the paths from RRT-connect by a simple and efficient method such that the difference is negligible

Exploit Morphology to Simplify Docking of Self-reconfigurable Robots

In this paper we demonstrate how to dock two self-reconfigurable robots and as a result merge them into one large robot. The novel feature of our approach is that the configuration we choose for our robots allows the robots to handle misalignment errors and dock simply by pushing against each other. In 90 experiments with the ATRON self-reconfigurable robot we demonstrate that two three-module robots can dock in 16 seconds without using sensors and are successful in between 93% and 40% of the attempts depending on approach angle and offset. While this is a modest step towards fast and reliable docking, we conclude that choosing appropriate configurations for docking is a significant tool for speeding up docking.

Neighbor detection and crosstalk elimination in self-reconfigurable robots

This paper addresses two issues concerning communication between neighbor modules in self-reconfigurable robots. The first issue is automatic neighbor detection that is due to modules self-reconfiguring, whereby the local communication network topology dynamically changes. The second issue is crosstalk between non-neighbor modules, where data packages send through an infrared communication channel are received by a non-neighbor module because of reflections. In this paper, we proposed algorithmic solutions to automatic neighbor detection and crosstalk elimination. The algorithms are simple, distributed, self-organizing and robust. For validation, they are implemented and evaluated on the physical ATRON system. In conclusion, the algorithms are efficient and effective and we argue that these algorithmic contributions may be applicable on other systems as well.

Efficient enumeration of modular robot configurations and shapes

A modular robot consists of a set of mechatronic modules that can be connected in many different ways, which makes it possible to build robots of many different shapes from the same basic set of modules. The main contribution of this work is an algorithm that, given the parameters of a module and the number of modules, efficiently can calculate how many different configurations and shapes can be built. These numbers are important because the first is a measure of the self-reconfigurability and the second, given there is a relationship between form and function, the versatility of a modular robot. As an experimental contribution, we enumerate the configuration and shape spaces of square, two-dimensional modules with all possible connector configurations. We proceed to three dimensions and enumerate the spaces of the theoretically interesting sliding cube module, and the M-TRAN and SuperBot modules. Several observations are made, an important one is that the shape spaces of the two physical modules are large even for a small number of modules (103 different shapes using 3 modules). This implies that it is not the lack of shape diversity that holds these modular robots back from being versatile. A result that suggests that if modules are designed right, versatility can potentially be reached even with few modules, which is contrary to the common belief in the community that more is better.

LocoKit: A Robot Construction Kit for Studying and Developing Functional Morphologies

We describe a robot construction kit named LocoKit and a method for studying functional morphologies. LocoKit consists of simple mechanical parts that allow for construction of a wide range of morphologies and modular electronics for instrumentation and control. The method relies on LocoKit for constructing functional morphologies and an experimental setup borrowed from the study of functional morphology in animals. We demonstrate the use of LocoKit and the method in a case study on quadruped locomotion and conclude that the methodology represents a systematic and efficient approach to the study and development of functional robot morphologies.

Increased performance in a bottom-up designed robot by experimentally guided redesign

Purpose – Using a bottom‐up, model‐free approach when building robots is often seen as a less scientific way, compared to a top‐down model‐based approach, because the results are not easily generalizable to other systems. The authors, however, hypothesize that this problem may be addressed by using solid experimental methods. The purpose of this paper is to show how well‐known experimental methods from bio‐mechanics are used to measure and locate weaknesses in a bottom‐up, model‐free implementation of a quadruped walker and come up with a better solution.Design/methodology/approach – To study the bottom‐up, mode‐free approach, the authors used the robotic construction kit, LocoKit. This construction kit allows researchers to construct legged robots, without having a mathematical model beforehand. The authors used no specific mathematical model to design the robot, but instead used intuition and took inspiration from biology. The results were afterwards compared with results gained from biology, to see if ...