A Novel Docking System for Modular Self-Reconfigurable Robots

In this thesis we present the design and implementation of a novel self-reconfiguring modular (SR-MR) robotic system: Roombots. We are aiming at three main applications with Roombots; locomotion through self-reconfiguration in the regular cubic 3D-lattice on structured surfaces, locomotion in non-structured environments applying central pattern generators (CPG) as the locomotion controller, and self-assembly and reconfiguration of static objects of the day-to-day environment, such as furniture. Robot assemblies from self-reconfigurable modular robots have the ability to adapt to a given task and working environment by altering their shape through a series of reconfiguration moves, and attachments and detachments between the modules. We are interested in self-reconfiguring modular robots for their shape-changing capabilities, and their distributed characteristics. We envision the following applications for Roombots: First, self-reconfiguration in a structured 3D lattice, i.e. a floor and walls equipped with connectors. Embedded connectors can provide pivot points for locomotion of SR-MR assemblies, and docking and recharging places for our adaptive furniture pieces. Second, our proposed concept for locomotion control of modular robots on non-structured ground are central pattern generators. Third, we would like to build adaptive and versatile furniture from modular robots and light-weight elements. In the following we are able to count more than 60 modular robotic systems, developed over the last two decades. However we were unable to identify an existing system which could provide us with all desired kinematic and geometric capabilities. This led us to design and implement a novel self-reconfiguring modular robotic system: Roombots. To tackle the module design we attempt to identify both meaningful design parameters from existing modular robots, and essential features for our applications. The combination of both leads to the kinematic and geometric description of the Roombots modules, and eventually to its implementation. In order to be able to assemble furniture from our Roombots units in the future, we need a reconfiguration framework which supports the specific requirements of Roombots. Metamodules made of two units attached in-series are attracted and guided by a virtual force-field, they use broadcast signals, look-up tables of collision clouds and simple assumptions about their near environment to reach their seeding positions, which are currently hand coded. For the task of locomotion in non-structured environments we propose a framework for learning to move with modular robots using central pattern generators and online optimization. The distributed implementation of CPGs offers an ideal substrate for producing locomotion patterns and for online learning, and an optimization framework for fast learning.

Self-reconfigurable robots are built from robotic modules typically organised in a lattice. The robotic modules themselves are complete, although simple, robots and have onboard batteries, actuators, sensors, processing power, and communication capabilities. The modules can automatically connect to and disconnect from neighbour modules and move around in the lattice of modules. The self-reconfigurable robot as a whole can, through this automatic rearrangement of modules, change its own shape to adapt to the environment or as a response to new tasks. Potential advantages of self-reconfigurable robots are extreme versatility and robustness. The organisation of self-reconfigurable robots in a lattice structure and the emphasis on local communication between modules mean that lattice automata are a useful basis for control of self-reconfigurable robots. However, there are significant differences which arise mainly from the physical nature of self-reconfigurable robots as opposed to the virtual nature of lattice automata. The problems resulting from these differences are mutual exclusion, handling motion constraints of modules, and unrealistic assumption about global, spatial orientation. Despite these problems the self-reconfigurable robot community has successfully applied lattice automata to simple control problems. However, for more complex problems hybrid solutions based on lattice automata and distributed algorithms are used. Hence, lattice automata have shown to have potential for the control of self-reconfigurable robots, but still a unifying implementation based on lattice automata solving a complex control problem running on physical self-reconfigurable robot is yet to be demonstrated.

Modular robot is an approach to build the robot for various complex tasks with the promise of great versatility, robustness and lower cost. It can be used extensively to meet the demands of different tasks or different working environments by changing its shapes. Therefore, they can travel over or through obstacles, and go though small pipe. Even they can walk somewhat like a person on crutches. It is a key technology that how modular robot with vision to capture exact image information, as well as to extract the feature parameters of components real-timely, to recognize the component types, and to judge the position and posture of component. The quantity of the motion pattern will decide the adjustable ability of modular robot. So, dynamic vision control system is very important for modular robot, which can improve its moving ability and intelligent degree. With the help of the dynamic vision control system, the robot can accomplish different tasks in different working environments by itself, such as deciding routes and avoiding obstacles. In this paper, in order to improve the vision of modular robot, a kind of dynamic vision control system is analyzed and researched roundly, and two kinds of motion patterns for the modular robot are put forward.

Self-reconfigurable robot is a kind of robot that is able to recover its original function after partial damage of the robotic system. This is achieved by making changes on the partially damaged robot. This paper put forward a kind of self-reconfigurable quadruped robot based on the resilient theory. First, the concept of the self-reconfigurable robot and the recovery strategy were introduced. Secondly, a novel architecture of modular quadruped robot made by LEGO material was proposed. The connector of interfaces was taken into consideration. At last, the overall mode of the motion and control method with the help of NXT G program were proposed. The analysis of the overall gait ensures the stability of motion in this phase. As for the uncertain environmental damage to the robot, the robot is able to repair itself by changing the movement type automatically to ensure it continues to work. The result shows that the quadruped robot meets the requirements of resilience and self-reconfiguration.

Modular self-reconfigurable robots (MSRs) are mostly used in environments where it is difficult to navigate and explore otherwise. Especially, the shape-changing ability of MSRs makes them more dexterous in these situations compared to fixed-body robots. But when the MSR forms a new configuration, usually, the locomotion pattern for that particular configuration is not known by its constituting robotic modules. The main challenge for modules is to learn how to move in that specific configuration within a reasonable amount of time. In this paper, we study the problem where an MSR needs to learn its movement pattern on-the-fly. To solve this problem, we have proposed a game theoretic solution based on multi-agent reinforcement learning using which the constituting modules distributedly learn the best actions that they need to perform to travel more distance in less time. We have implemented this approach in simulation on both ModRED and Yamor MSR platforms. Results show that our approach performs better (up to 7.86 times) in terms of average speed achieved for most of the tested configurations as compared to an existing locomotion learning approach.

Self-reconfigurable robots are constructed of robotic modules that can be connected in many different ways. These modules move in relationship to each other, which allows the robot as a whole to change shape. This shapeshifting makes it possible for the robots to adapt and optimize their shapes for different tasks. Thus, a self-reconfigurable robot can first assume the shape of a rolling track to cover distance quickly, then the shape of a snake to explore a narrow space, and finally the shape of a hexapod to carry an artifact back to the starting point. The field of self-reconfigurable robots has seen significant progress over the last twenty years, and this book collects and synthesizes existing research previously only available in widely scattered individual papers, offering an accessible guide to the latest information on self-reconfigurable robots for researchers and students interested in the field. Self-Reconfigurable Robots focuses on conveying the intuition behind the design and control of self-reconfigurable robots rather than technical details. Suggestions for further reading refer readers to the underlying sources of technical information. The book includes descriptions of existing robots and a brief history of the field; discussion of module design considerations, including module geometry, connector design, and computing and communication infrastructure; an in-depth presentation of strategies for controlling self-reconfiguration and locomotion; and exploration of future research challenges.

In this paper, a novel robotic system called modular transformer (M-TRAN) is proposed. M-TRAN is a distributed, self-reconfigurable system composed of homogeneous robotic modules. The system can change its configuration by changing each module's position and connection. Each module is equipped with an onboard microprocessor, actuators, intermodule communication/power transmission devices and intermodule connection mechanisms. The special design of M-TRAN module realizes both reliable and quick self-reconfiguration and versatile robotic motion. For instance, M-TRAN is able to metamorphose into robotic configurations such as a legged machine and hereby generate coordinated walking motion without any human intervention. An actual system with ten modules was built and basic operations of self-reconfiguration and motion generation were examined through experiments. A series of software programs has also been developed to drive M-TRAN hardware, including a simulator of M-TRAN kinematics, a user interface to design appropriate configurations and motion sequences for given tasks, and an automatic motion planner for a regular cluster of M-TRAN modules. These software programs are integrated into the M-TRAN system supervised by a host computer. Several demonstrations have proven its capability as a self-reconfigurable robot.

The advantages of modular robot systems stem from their ability to change between different configurations, enabling them to adapt to complex and dynamic real-world environments. Then, how to perform the accurate and efficient change of the modular robot system, i.e., the self-reconfiguration problem, is essential. Existing reconfiguration algorithms are based on discrete motion primitives and are suitable for lattice-type modular robots. The modules of freeform modular robots are connected without alignment, and the motion space is continuous. It renders existing reconfiguration methods infeasible. In this paper, we design a parallel distributed self-reconfiguration algorithm for freeform modular robots based on multi-agent reinforcement learning to realize the automatic design of conflict-free reconfiguration controllers in continuous action spaces. To avoid conflicts, we incorporate a collaborative mechanism into reinforcement learning. Furthermore, we design the distributed termination criteria to achieve timely termination in the presence of limited communication and local observability. When compared to the baselines, simulations show that the proposed method improves efficiency and congruence, and module movement demonstrates altruism.

A modular robot has a distributed mechanical composition which can make various configurations and also make locomotion in a wide variety of configurations. Modular robots are thought to be useful in extreme or unknown environments by adaptively changing their shape and locomotion patterns. As for locomotion, two types can be used; one is whole-body fixed-configuration locomotion and the other is locomotion by self-reconfiguration. In this paper we deal with the former type of locomotion which is realized by coordinated joint actuation. So far, proposed control methods for whole-body locomotion by modular robots have been based on predefined locomotion sequences. However, locomotion based on predefined sequences cannot adapt to changing terrain conditions such as uphill, downhill, slippery and sticky grounds. To solve such problems, we propose a distributed control mechanism using a CPG controller which enables adaptive locomotion by modular robots. Besides the real-time CPG control we introduce a decentralized control mechanism for detecting the situation that the robot is stuck and initiating transformation to another shape for recovering the situation. The results of various hardware experiments by 4-legged structure prove the feasibility of the method for adaptive locomotion and transformation by our M-TRAN II modules.

Modular robotics is a popular topic for robotic applications and design. The reason behind this popularity is the ability to use and reuse the same robot modules for accomplishing different tasks through reconfiguration. The robots are capable of self-reconfiguration based on the requirements of the task and environmental constraints. It is possible to have a large number of configuration combinations for the same set of modules. Therefore, it is important to identify unique configurations from among the full set of possible configurations and establish a kinematic strategy for each before reconfiguring the robots into a new shape. This becomes more difficult for robot units having more than one connection type and more degrees of freedom (DOF) For example, ModRED II modules have two types of connections and four DOF per module. In this paper, the set of configurations is enumerated, and determination of configuration isomorphism is accomplished for ModRED II modules using graph theory. Kinematic equations are then derived for unique configurations. The kinematic method is then demonstrated for certain example configurations using ModRED II modules.

This paper presents an automatic coupling mechanism with mechanical overload protection for use in modular robotic systems. The genderless coupling incorporates a novel safety mechanism that disengages two modules at a predefined torque threshold. Several key design aspects allow the minimalist coupling to contain all functionalities within a compact cylindrical space. A detailed analytical model of the overload protection mechanism is validated by testing two materials and fabrication methods. We further present a functional prototype of the coupling with a demonstration of two modular robots driving toward one another, automatically connecting and disconnecting, detecting the connection, communicating in the process, and sensing the angle between them. The coupling as well as the embedded technologies and mechanisms present essential steps toward enabling fully self-reconfigurable robots combining modular and origami features, while advancing the process of introducing modularity into a wide range of robotic systems.

Compared with the fixed morphological robots, the self-reconfigurable robots (SR) are very flexible as they can adapt to a wide range of task requirements. However, the current efforts on SR mainly focus on the locomotion issues, while the manipulation capabilities of existing systems are still quite limited. In this paper, a new kind of mobility reconfigurable robotic structures (MRRS) with self-reconfiguration ability is presented. The reconfigurable topology and variable mobility characteristics of the MRRS are realized by integrating the cooperative kinematic mechanism (CKM) and the separable parallel kinematic mechanism (SPKM) into one robotic system, which makes the robot capable of reconfiguring its mechanical assembly to adapt to the multiple motion task requirements. Based on Lie group of displacements, the commonly used robotic tasks are first classified into three categories: positioning motion, tooling motion and grasping motion, possible motion types of these tasks are enumerated, then how to deal with the multiple tasks by using the MRRS is synthesized systematically. Finally, a multi-task working process is employed as an example to illustrate the advantages of this reconfigurable robotic structure.

Graph theory plays a pivotal role in the self-reconfiguration strategies of robots, optimizing the strategies of Modular Self-Reconfigurable Robots (MSRR) in unknown environments. By incorporating concepts from graph theory, particularly adjacency matrices and Voronoi diagrams, a novel graph-based strategic approach can be devised to effectively guide robots through self-reconfiguration and enhance their adaptability to the environment. Initially, adjacency matrices are utilized to describe the interconnections between robot modules, capturing the robot system's topological structure with precision. Subsequently, the application of Voronoi diagrams further bolsters the understanding of robots’ relative spatial positioning and adjacency, providing vital spatial information for autonomous exploration and decision-making. The synergy of these graph theory tools not only amplifies the analysis of environmental adaptability but also furnishes new theoretical support and practical pathways for strategy selection during the robots’ self-reconfiguration process. In practice, the potential energy values of graphs can be used to illustrate and optimize the robots’ behavioral strategies, revealing the practical value of graph theory in robotic strategy optimization and enabling more accurate adaptation to complex and dynamic environmental conditions. This graph theory-based strategy optimization method not only increases the computational efficiency and decision quality of the robot system but also offers solutions to new challenges encountered in robot technology.

Modular robot systems towards the execution of cooperative tasks in large facilities

Transferring motion instructions from a user enables robots to perform new and unforeseen operations. Robot collectives, in particular, offer greater adaptability to changing tasks and environments. However, effectively transferring motion instructions becomes challenging as the collective's shape and size evolve. These changes often require additional system constraints to maintain robust motion control, which typically depends on pre-programmed knowledge of new tasks, ultimately limiting the collective's adaptability. To overcome the above challenges, we propose a physical and computational platform for user-guided control of self-reconfigurable modular robots. This platform consists of an optimization scheme for online processing of user commands, which prevents any modular robot actions that would violate system or environment constraints. The second component consists of Joint-space Joysticks, which match the robot's morphology, enabling the user to control diverse and dynamically changing modular robot structures through direct physical interaction. We present a platform that enables users to safely control modular, shape-changing robots through a physical interface. We demonstrate the platform's efficacy and generalizability across a diverse set of modular robot morphologies using two independent robotic systems - Mori3 and Roombots - performing a range of tasks including pick-and-place, human assistance, legged locomotion, and workspace expansion.