Akos Csiszar

EcoBotics: Advantages and challenges of building a bamboo robot arm

Sustainability and sustainable development concepts are frequently the driving force for research related to optimizing energy consumption and material use in order to reduce the carbon footprint for different products. EcoBotics aims to analyse the question of sustainability in the area of robotics. In this paper, as a first example of EcoBotics, it is examined what are the advantages and challenges of exchanging the aluminium structure of lightweight robot arms with a sustainable material. Bamboo is used in order to exemplify a sustainable material, but the challenges and the proposed solutions for open problems are valid for any material causing a robot structure with low stiffness.

A combinatorial optimization approach to the Denavit-Hartenberg parameter assignment

Assigning a Denavit Hartenberg parameter table to a serial robot or a kinematic chain is a cumbersome task which confuses many novice roboticists. In this paper a combinatorial optimization approach to Denavit Hartenberg parameter assignment is proposed. The proposed combinatorial optimization approach eliminates the reliance on human experience when assigning Denavit-Hartenberg parameters. Using practical insights the parameter assignment problem is transferred from a continuous search space to a discrete search space hence enabling the use of combinatorial optimization techniques. The search space is reduced below a limit which makes the application of exhaustive search and branch and bound optimization techniques practicable. An algorithm is described which is capable of generating all practical Denavit Hartenberg parametrizations of a kinematic chain (including dummy frames) within an acceptable time frame. The obtained results are ranked based on a proposed criterion.

Optimal Design Concept of a Reconfigurable Haptic Device

Haptic devices are more and more common for human computer interaction or human machine interaction type applications. Using such devices multimodal interaction can be realized, increasing this way immersibility in the virtual reality environment. This paper presents specific considerations regarding the optimal design process of a class of reconfigurable haptic devices. The main focus of the paper is on the presentation of the novel ReHapy concept and on the details regarding the formulation of the cost function on which the optimal design process will be based.

Fast robot task and path planning based on CAD and vision data

Real manufacturing environments are cluttered spaces, where the environment matches its digital model (the CAD model) only to a certain degree. Robot systems, which autonomously execute tasks in these environments should combine CAD and vision data in order to successfully carry out their tasks. This way the model data (coming from the CAD model) can be corrected by the sensor data (coming from computer vision) to properly reflect the real environment. In this paper, a novel method for autonomous task-space exploration is presented, which is based on space exploration and object recognition. During the exploration of the environment, the CAD data is used to determine the initial belief of the newly gained information. After constructing the environment model, a task planner based on a geometric contact analysis and symbolic inference computes the necessary manipulation sequence. To compute the disassembly space, a novel time-efficient algorithm is proposed. The theoretical aspects are applied to disassembly sequences. A variety of heuristics for computing a local optimal sequence of the required manipulation steps and the corresponding paths is also presented in the paper. To plan the paths well-known global path planners are applied with a step-size control which reduce the planning effort. The approaches are experimentally validated and the results are discussed.

Behavior Trees for Task-Level Programming of Industrial Robots

The number of industrial robots used worldwide has been continuously increasing. In almost all cases the application program has the form of a text based source code, which has inherent drawbacks (in terms of complexity and ease of development) when compared to the graphical approaches seen in general purpose software engineering. These graphical programming approaches have not been developed, having industrial robots in mind. In this paper a behavior trees based approach for creating and representing source code for robotic applications is proposed. The software architecture and an experimental implementation of the developed programming method is presented and a validation using a common assembly task is shown.

A Concept for the Application of Reinforcement Learning in the Optimization of CAM-Generated Tool Paths

Cyber physical systems (CPS) are changing the way machine tools function and operate. As the CAD-CAM-CNC tool chain gains intelligence the boundaries of the elements of the tool chain become blurred and new features, based on advancements in artificial intelligence can be integrated. The main task of the CAD-CAM-CNC chain is to generate the cutter trajectories for the manufacturing operation. Driven by sustainability and the need for capacity, the need arises to optimize the paths through this tool chain. In this paper a concept for path optimization with reinforcement learning is proposed, with focus on the reward function, specific to tool path optimization via the channel method.

A combinatorial approach to the automated generation of inverse kinematics equations for robot arms

Obtaining the inverse kinematics equations (incl. the higher order ones) is an essential part of most modeling tasks related to robotics. It should be routine task which in practice proves to be difficult because of its error prone nature. In this paper an automated method of obtaining these equations is proposed. The mathematical operations which are carried out also when the equations are written by hand, are formulated as possible actions to take. A correct sequence of such actions leads to the desired equations. This way the problem is formulated as a planning problem. In this paper, as a preliminary approach, the breadth first search algorithm is used to solve the planning problem, but this formulation of the problem enables also the use of reinforcement learning methods. An implementation well suited for further expansions and source code generation for multiple programming languages is also presented.

Conceptual development of a reconfigurable parallel robot

Todays rapidly changing market demands and competitive industry requires timely reactions to market changes. The usage of reconfigurable systems permits a fast reaction to changes in a cost-effective manner. In this paper a new class of parallel reconfigurable robots is introduced. Configurations differ in their number of degrees of freedom. Aspects of topological reconfiguration are presented. Virtual models of the robots have been developed and presented in this paper for an early validation of theoretical consideration. Using these virtual models, the motion capabilities of the various configurations have been analyzed. Numerical results are presented at the end of the paper.