Efficient Evolution of Modular Robot Control via Genetic Programming

In Genetic Programming (GP), most often the search space grows in a greater than linear fashion as the number of tasks required to be accomplished increases. This is a cause for one of the greatest problems in Evolutionary Computation (EC): scalability. The aim of the work presented here is to facilitate the evolution of control systems for complex robotic systems. The authors use a combination of mechanisms specifically designed to facilitate the fast evolution of systems with multiple objectives. These mechanisms are: a genetic transposition inspired seeding, a strongly-typed crossover, and a multiobjective optimization. The authors demonstrate that, when used together, these mechanisms not only improve the performance of GP but also the reliability of the final designs. They investigate the effect of the aforementioned mechanisms on the efficiency of GP employed for the coevolution of locomotion gaits and sensing of a simulated snake-like robot (Snakebot). Experimental results show that the mechanisms set forth contribute to significant increase in the efficiency of the evolution of fast moving and sensing Snakebots as well as the robustness of the final designs.

Design for Information Processing in Living Neuronal Networks

A neurorobot is a model system for biological information processing with vital components and the artificial peripheral system. As a central processing unit of the neurorobot, a dissociated culture system possesses a simple and functional network comparing to a whole brain; thus, it is suitable for exploration of spatiotemporal dynamics of electrical activity of a neuronal circuit. The behavior of the neurorobot is determined by the response pattern of neuronal electrical activity evoked by a current stimulation from outer world. “Certain premise rules” should be embedded in the relationship between spatiotemporal activity of neurons and intended behavior. As a strategy for embedding premise rules, two ideas are proposed. The first is “shaping,” by which a neuronal circuit is trained to deliver a desired output. Shaping strategy presumes that meaningful behavior requires manipulation of the living neuronal network. The second strategy is “coordinating.” A living neuronal circuit is regarded as the central processing unit of the neurorobot. Instinctive behavior is provided as premise control rules, which are embedded into the relationship between the living neuronal network and robot. The direction of self-tuning process of neurons is not always suitable for desired behavior of the neurorobot, so the interface between neurons and robot should be designed so as to make the direction of self-tuning process of the neuronal network correspond with desired behavior of the robot. Details of these strategies and concrete designs of the interface between neurons and robot are be introduced and discussed in this chapter.

Gait Transition Control of a Biped Robot from Quadrupedal to Bipedal Locomotion Based on Central Pattern Generator, Phase Resetting, and Kinematic Synergy

Recently, interest in the study of legged robots has increased, and various gait patterns of the robots have been established. However, unlike humans and animals, these robots still have difficulties in achieving adaptive locomotion, and a huge gap remains between them. This chapter deals with the gait transition of a biped robot from quadrupedal to bipedal locomotion. This gait transition requires drastic changes in the robot posture and the reduction of the number of supporting limbs, so the stability greatly changes during the transition. A locomotion control system is designed to achieve the gait transition based on the physiological concepts of central pattern generator, phase resetting, and kinematic synergy, and the usefulness of this control system is verified by the robot experiment.

Robotic CAM System Available for Both CL Data and NC Data

This chapter describes the development of a robotic CAM system for an articulated industrial robot from the viewpoint of robotic servo controller. It is defined here that the CAM system includes an important function that allows an industrial robot to move along not only numerical control data (NC data) but also cutter location data (CL data) consisting of position and orientation components. A reverse post-processor is proposed for the robotic CAM system to online generate CL data from the NC data generated for a five-axis NC machine tool with a tilting head, and the transformation accuracy about orientation components in CL data is briefly evaluated. The developed CAM system has a high applicability to other industrial robots with an open architecture controller whose servo system is technically opened to end-users, and also works as a straightforward interface between a general CAD/CAM system and an industrial robot. The basic design of the robotic CAM system and the experimental result are presented, in which an industrial robot can move based on not only CL data but also NC data without any teaching.

Design and Development of Teleoperation for Forest Machines

Teleoperation of forestry machinery is a difficult problem. The difficulties arise because forestry machines are primarily used in unstructured and uncontrolled environments. However, improvements in technology are making implementation of teleoperation for forestry machines feasible with off-the-shelf computing and networking hardware. The state-of-the-art in teleoperation of forestry machinery is reviewed as well as teleoperation in similarly unstructured and uncontrolled environments such as mining and underwater. Haptic feedback in a general sense is also reviewed, as while haptic feedback has been implemented on some types of heavy machinery it has not yet been implemented on forestry machinery.

Development and Simulation of an Adaptive Control System for the Teleoperation of Medical Robots

This chapter presents the design and calculation procedure for a teleoperation and remote control of a medical robot that can help a doctor to use his hands/fingers to examine patients in remote areas. This teleoperation system is simple and low cost, connected to the global Internet system, and through the interaction with the master device, the medical doctor is able to communicate control signals for the slave device. This controller is robust to the time-variant delays and the environment uncertainties while assuring the stability and the high transparent performance. A novel theoretical framework and algorithms are developed with time forward observer-based adaptive controller and neural network-based multiple model. The system allows the medical doctor to feel the real sense of the remote environments.

Smart Sensor Systems

This chapter addresses smart sensor systems. In recent years, goods identification technology using a soft magnetic barcode, radio frequency identification, and automated wheelchair guidance technology using a magnetic field usable in dirty environments as part of Robotics and Mechatronics are becoming important in many areas, such as factories, physical distribution, office, security, etc. These identification and guidance technologies are based on sensing of magnetic field. Therefore, smart magnetic sensing technologies suitable for these identification and guidance techniques are described in this chapter.

Ellipse Detection-Based Bin-Picking Visual Servoing System

In this chapter, the authors tackle the task of picking parts from a bin (bin-picking task), employing a 6-DOF manipulator on which a single hand-eye camera is mounted. The parts are some cylinders randomly stacked in the bin. A Quasi-Random Sample Consensus (Quasi-RANSAC) ellipse detection algorithm is developed to recognize the target objects. Then the detected targets’ position and posture are estimated utilizing camera’s pin-hole model in conjunction with target’s geometric model. After that, the target, which is the easiest one to pick for the manipulator, is selected from multi-detected results and tracked while the manipulator approaches it along a collision-free path, which is calculated in work space. At last, the detection accuracy and run-time performance of the Quasi-RANSAC algorithm is presented, and the final position of the end-effecter is measured to describe the accuracy of the proposed bin-picking visual servoing system.

Silicon Micro-Robot with Neural Networks

Insect type 4.0, 2.7, 2.5 mm. width, length, height size silicon micro-robot system with active hardware neural networks locomotion controlling system is presented in this chapter. The micro-robot system was made from a silicon wafer fabricated by Micro-Electro Mechanical Systems (MEMS) technology. The mechanical system of the robot equipped with millimeter-size rotary type actuators, link mechanisms, and six legs to realize the insect-like switching behavior. In addition, the authors constructed the active hardware neural networks by analog CMOS circuits as a locomotion controlling system. Hardware neural networks consisted of pulse-type hardware neuron models as basic components. Pulse-type hardware neuron model has same basic features of biological neurons such as threshold, refractory period, spatio-temporal summation characteristics, and enables the generation of continuous action potentials. The hardware neural networks output the driving pulses using synchronization phenomena such as biological neural networks. Four output signal ports are extracted from hardware neural networks, and they are connected to the actuators. The driving pulses can operate the actuators of silicon micro-robot directly. Therefore, the hardware neural networks realize the robot control without using any software programs or A/D converters. The micro-robot emulates the locomotion method and the neural networks of an insect with rotary type actuators, link mechanisms, and hardware neural networks. The micro-robot performs forward and backward locomotion, and also changes direction by inputting an external trigger pulse. The locomotion speed was 26.4 mm/min when the step width was 0.88 mm.

Cyberinfra Product Concept and its Prototyping Strategies

The Knowledge-Intensive Sustainable Evolution Dynamics (KISBED) (patent pending), a platform the authors use in their “use-cases,” shows that it works. Cyber, infrastructure, and product are integrated in the Cyberinfra Product “function.” The perception properties are not long tagged or have no carriers, and the signal travels a short distance before it collides. The authors prove the KISBED through some examples.