Robust Robot Research Articles

Image segmentation techniques and their use in artificial life robot implementation

In this article we give a brief description of the properties of human and robot vision systems. We focus on the part of vision processing that is connected with the image segmentation problem, and especially with texturebased segmentation. Such vision processing is necessary for robust indoor mobile robot navigation and for other related tasks, including line-following, and object detection and recognition. We examine the details of a texture segmentation technique based on the Gabor filtering method.

Parameter estimation and upper bounding adaptation in adaptive-robust control approaches for trajectory control of robots

In this paper a new robust adaptive control law for n-link robot manipulators with parametic uncertainties is derived using the Lyapunov theory thus guaranteed the stability of an uncertain system. The novelty of the adaptive robust control algorithm is that manipulator parameters and adaptive upper bounding functions are estimated to control the system properly, and the adaptive robust control law is also updated as an exponential function of manipulator kinematics, inertia parameters and tracking errors. The proposed adaptive control input includes a parameter estimation law as an adaptive controller and an additional control input vector as a robust controller. The developed approach has the advantages of both adaptive and robust control laws, without their discolour tags.

Sensor information space for robust mobile robot path planning

This paper introduces a new approach to robust path planning for mobile robots entirely based on information from environment perception sensors. This method avoids the use of odometry which leads to the accumulation of errors resulting from the robot's position computing. We proceed as follows: we create regions inside which the robot detects the same obstacle segments. A node graph represents all the regions and their links. Then a planning algorithm is used to find a path which joins a start to a goal region. The final stage consists in applying a robust robot motion control as regards the uncertainties of the environment model. This approach contributes to a control system for indoor robots which is environment referenced. The sensors we deal with are first a continuous laser or ultrasonic scanning system, then a discrete ultrasonic belt whose limits of use we show.

Robust neural force control scheme under uncertainties in robot dynamics and unknown environment

The original impedance function is known to lack robustness due to unknown robot dynamic model and the environment. In order to improve that result, a new impedance function is derived which specifies a desired force directly. This results in a new robust robot force tracking impedance control scheme, which employs a neural network as a compensator to cancel out all uncertainties. The proposed neural force control scheme is capable of making the robot track a specified desired force as well as of compensating for uncertainties in environment location and stiffness, and in robot dynamics. Separate training signals for free-space motion and contact-space motion control are developed to train the neural compensator online. The design of the training signals is justified. Simulation studies with a three-link rotary robot manipulator are carried out and the results show excellent force tracking performance.

大型ピストン加工用ＦＭＳの構築

The introduction of FMS (Flexible Manufacturing System) in machining of large-size pistons (greater than outer diameter 170 mm) has been desired, because various shapes are demanded and moreover the production lot size is small. The automation in FMS construction is essential, but in large-size pistons this is very difficult because the following problems have to be overcome: (1) The integration of various operations for effective floor space utilization. (2) The robot loader with the ability of accurate positioning as well as heavy movable load carrying capacity is demanded, because the positioning accuracy is necessary to the same degree of small (or middle)-size piston. (3) Measurement after every operation is essential to prevent the production of defective pistons, because the unit cost of the large-size piston is very high. (4) The dimension of the outer diameter in the machining of large-size pistons should be automatically controlled, but it has been dependent on the worker's skill owing to the complicated shape and its variety. The FMS construction for large-size piston machining was completed by applying the following representative techniques: (1) Integrated system of multiple operations for effective floor space utilization. (2) Automatic conveyor system consisting of robust robot loaders with accurate positioning and AGV (Automated Guided Vehicle). (3) Inner cooling pipe sensing by the ultrasonic inspection instrument and measuring system. (4) Development of non-circular outer diameter finishing machine being capable of machining complicated profiles and any type of piston. The advantage of constructed FMS is that the productivity increases to about 3 times as compared to the conventional manual line.

Communication and propagation of action knowledge in multi-agent systems

Multi-agent systems (MASs) do play an important role in the construction of fault tolerant and robust robot systems. One major advantage of MAS is the fact that multiple agents work towards a common goal, having different skills for specific subtasks. Usually, agents have to use a common description of the actions to be carried out. Since agents can join and leave the MAS at any time, it is important that knowledge acquired by single agents can be transferred or propagated between agents, to ensure that knowledge is not lost, in the case that an agent leaves the system. In this paper, techniques will be presented that enable representation of common extendable action knowledge for task solutions in an agent’s knowledge base and additionally, algorithms for propagating this knowledge between agents efficiently and with minimum required communication effort.

Automatic Mapping of Dynamic Office Environments

We present a robot, InductoBeast, that greets a new office building by learning the floorplan automatically, with minimal human intervention and a priori knowledge. Our robot architecture is unique because it combines aspects of both abductive and inductive mapping methods to solve this problem. We present experimental results spanning three ofiice environments, mapped and navigated during normal business hours. We hope these results help to establish a performance benchmark against which robust and adaptive mapping robots of the future may be measured.

Robot vision robust to changes of lighting conditions using a wide-dynamic-range visual sensor

We have developed a robust vision system for 2D positioning of industrial parts stable to changes of lighting conditions. A wide dynamic range vision sensor that we had developed previously was used to avoid the saturation of object images. Additionally, a gray scale pattern matching technique was employed for robust image processing. Performance of the system with a wide dynamic range vision sensor was investigated experimentally in comparison with that of a system with a conventional video camera. The probability of correct positioning of an object under changing lighting conditions, simulating those in a factory, was 100% for the wide dynamic range vision sensor and 83% for the conventional video camera. The experimental results confirm the effectiveness of the developed system, revealing that dynamic range expansion of the video cameras is very effective for realizing robust robot vision systems. © 1997 Scripta Technica, Inc. Electr Eng Jpn, 120(2): 34–40, 1997

Motion of control of a multijoint robot based on a robust velocity controller

This paper proposes a simple and robust robot motion control method using a robust velocity controller. The robust velocity controller is based on H∞ control theory, and is called H∞ velocity controller. The H∞ velocity controller based motion control method is completely equivalent to the robust acceleration control method using the H∞ acceleration controller, but it has simpler structure. Therefore, the proposed system can realize a fine robot motion control easily. To confirm the validity of the proposed method, this paper realizes the hybrid control of position and force for a multijoint robot manipulator. Further, the simple realization of hybrid control is proposed considering the attitude of the robot manipulator. This system achieves hybrid control of position and force of the robot manipulator while maintaining a perpendicular attitude to the target environment. The experimental results in this paper show that the proposed system has the desired force and position response to the target environment. © 1997 Scripta Technica, Inc. Electr Eng Jpn, 118 (4): 58–69, 1997

Robust robot manipulator control with autonomous consideration algorithm of torque saturation

As each joint actuator of a robot manipulator has a limit value of torque, the motion control system should consider the torque saturation. Conventional motion control based on robust acceleration controller cannot consider the torque saturation and it often causes an oscillated or wrong response. This paper proposes a new autonomous consideration method of joint torque saturation for robust manipulator motion control. The proposed method consists of three on-line autonomous algorithms. These algorithms are the torque limitation algorithm in joint space, the adjustment algorithm of motion control in Cartesian space, and the adjustment algorithm of motion reference in Cartesian space. The robot motion control using the proposed algorithms realizes smooth and robust robot motion response.

Detecting and diagnosing mistakes in vision-based navigation

Abstract Robust mobile robot navigation is an open problem. Navigational mistakes are inevitable due to the characteristics of sensors, models, actuators, and natural environments. This paper describes how to detect and diagnose mistakes that autonomous mobile robots make while navigating through large-scale space using vision. Mistakes are perceptual, cognitive, or motor events that lead a robot astray from its intended route. Detecting and diagnosing a mistake involves realizing that this has happened and determining what the mistake was and when it happened. The approach described in this paper handles mistakes explicitly. Mistakes are detected by looking for mismatches between observations and expectations. Detailed symbolic representations of visual information support comparing observations and expectations augmented by a priori knowledge. Mistakes are diagnosed by examining knowledge from a variety of sources, including a history of the mobile robot's observations and actions. A computer program called Muckle implements this approach in simulation and provides experimental results that demonstrate the feasibility and potential of the approach.

照明変動にロバストな部品位置決め用ロボット視覚システム

A robust vision system against the change of lighting condition for industrial parts positioning has been developed. A wide dynamic range vision sensor, we had developed previously, was employed to avoid the saturation of images of the objects. Performance of the developed system with the wide dynamic range vision sensor was investigated experimentally in comparison with that of the system with a conventional video camera. The experimental result has confirmed the effectiveness of the developed system, revealing that the dynamic range expansion of the video cameras is very effective for realizing robust robot vision systems.

Vision for Space Robots on a Parallel Distributed Architecture

This article reports on two main contributions: the introduction of a parallel distributed architecture for robot vision, and robust robot vision techniques and their computationally efficient parallel implementation for grasping a polyhedral object

Extraction and computation of identifiable parameters in robot dynamic models: theory and application

The accuracy of robot dynamic models plays an important role in the design of stable and robust robot controllers. The accuracy of these dynamic models is largely dependent on the precise measurement of robot dynamic parameters (e.g first moments, radius of gyration matrix, etc.). Towards this end, algorithms for the identification of these parameters need to be developed to obtain dynamic models with high fidelity, which will ultimately improve the future designs of robot controllers. This work aims to provide a viable solution to this problem. A simplified model of the dynamics is used to extract and group the dynamic parameters of a robot manipulator. The resulting equations are mapped onto a network of parallel processors to speed up the rate of computations and enable real-time implementations. The efficiency of the algorithm is demonstrated by a case study.

Control of Nonlinear Systems Using Terminal Sliding Modes

Many robotic systems would, in the future, be reuqired to operate in environments that are highly unstructured and active, i.e., possessing means of self-actuation. Although a significant volume of results exist in model-based, robust and adaptive control literature, general issues pertinent to the performance of such control systems remain unresolved, e.g., feasibility of implementing high gain switches for control robustness. It is also pointed out that in certain applications, control switching can be very detrimental to the overall system. The primary focus of this paper is development of a new approach to control synthesis for robust robot operations in unstructured environments. To enhance control performance with full model information, we introduce the notion of terminal convergence, and develop control laws based upon a new class of sliding modes, denoted terminal sliders. We demonstrate that terminal sliders provide robustness to parametric uncertainty without having to resort to high frequency control switching, as in the case of conventional sliders [2]. In addition, stability analysis that is conducted to demonstrate terminal slider approach results in improved control performance and allows for simple robust design of control parameters. Further, improved (guaranteed) precision of terminal sliders is argued for through an analysis of steady state behavior.

Strictly positive real admittances for coupled stability

This paper builds upon recent work that has addressed the stability of a feedback- controlled robot coupled to a passive, dynamic environment. A new definition of a “strictly positive real” function is presented, and is used to provide necessary and sufficient conditions for the exponential stability of a coupled system comprising a 1-port robot with a strictly positive real admittance, and a 1-port environment with a positive real (but otherwise arbitrary) impedance. The distinction between the new definition and conventional definitions of a strictly positive real function is founded in physical systems theory; the new definition relies upon distinct roles for efforts and flows, and upon the concept of an excess state. This definition will provide a useful design constraint for the development of robust robot controllers.

MOBILE ROBOTS

Mobile robots must deal with unstructured environments in order to be general and useful. Current laboratory robots often use single approach to deal with unstructured environments: position sensing, environment sensing, planning, or mechanical design. Robust robots of the future will have to use combinations of all those approaches, and also handle the difficult problems of system integration. This article is an overview of the problems to be faced and of some of the more promising approaches being investigated.

Planning of sensing tasks in an assembly environment

The purpose of this paper is to give an overview of past and recent work on planning sensing strategies for vision sensors. To achieve an economic use of robots in manufacturing, their programs must provide a high degree of fault-tolerance, security, and robustness to prevent unforeseen errors. Model errors (also termed uncertainties) are one of the most frequent reasons for such undesirable events. Robot systems can be made more reliable and fault-tolerant by providing them with capabilities of error detection and recovery, or error prevention. The latter may be achieved by reducing model errors using tactile and non-tactile sensors. The quality of a robot program synthesized by a task-level programming system depends on the accuracy of the model, since all information that is not explicitly given by the programmer must be derived from it. This means that the following questions have to be answered by the automatic task planner in order to plan non-tactile sensing strategies: (1) When do I have to use sensors to reduce uncertainty about the real world? (2) What do I have to use them for? (3) How do I have to use them to achieve the necessary information within an acceptable period of time? There are very few systems which deal broadly with the problem of robust robot programs, whereas there are numerous works on detail aspects of the field. The main approaches will be introduced and discussed in more detail. Finally, a new concept for generating sensor-integrated robust robot programs will be proposed.

Robust Control of Robots with Joint Elasticities

Robust stability of robot manipulator control loops is investigated, where the class of robot models under consideration have flexible joints. In addition, the state-of-the-art of control techniques for robots with elastic joints is briefly summarized. For the robustness evaluation, a method which has previously been applied to rigid robot models is carried over to the elastic case, and some selected problems are pointed out. In a case study, the feasibility of the proposed robust robot control-loop design technique is discussed.

A vision system for the man on the shopfloor

ASEA, the Swedish robot builder, has introduced a robust robot vision system which is easy to program by the person on the shopfloor. John Mortimer reports.