Ralph O. Buchal

Implementation of the principal-axis method for machining of complex surfaces

In this paper a new strategy for 5-axis machining of complex surfaces is presented. The method uses curvature alignment and matching between the design surface and the cutting tool to improve surface finish and reduce machining time. The method is implemented on two configurations of 5-axis machines, and used to machine a test surface. The results of the tests show a considerable improvement over conventional 3-axis machining.

Cognitive factors in distributed design

Engineering design is being carried out by distributed design teams in an effort to use expert human resources more efficiently. The support needs of distributed design are reviewed from a cognitive viewpoint using five broad categories: design methodology, collaboration, teamwork, knowledge management and design representation. Web-based implementation of iterative, structured design methodologies allow greater accessibility to tools and methodologies and more consistent interfaces. Collaboration requires successful and efficient sharing of knowledge, negotiation, coordination and management of activities. In distributed environments, organizational factors and decisions that foster teamwork must be mediated by technology. Design intent, rationale and history are important basic types of knowledge that knowledge management systems are required to capture, organize and manipulate to help generate new design knowledge. Efficient methods of representing design artefacts in different forms are needed that allow designers to interact most efficiently as well as support knowledge capture, transformation and collaborative activities.

Simulated Off-Line Programming of Welding Robots

The objectives of the work described in this paper are to develop and evaluate a method for calculation of interference between objects in a robot welding workstation, to develop a robot trajectory planning technique recognizing interference, inverse kinematics constraints, and welding parameter re quirements, and to demonstrate the trajectory planning tech nique using advanced computer graphics. Workstation mod eling and relevant kinematic relationships are discussed. A method of interference calculation is developed which allows rapid calculation of interference between convex polyhedra using a steepest descent graph search. If desired, exact sepa rating distances and directions can be found. A simple mod ule for automatic selection of welding parameters is de scribed. Interference-free trajectories that do not violate kinematics constraints are found by searching among possi ble inverse kinematic solutions for the desired welding torch trajectory. The above methods have been combined and implemented using interactive graphics in a prototype soft ware package called A UTO WELD.

Sketching and computer-aided conceptual design

Sketching is widely considered to be an essential activity during conceptual design, and many argue that CAD tools should be faithful to the sketching metaphor for conceptual design. However, CAD tools have progressed significantly in recent years, and there is growing experimental evidence that existing CAD tools can be as effective as sketching. Recent research in cognitive psychology supports the idea that the sketching metaphor is not necessarily ideal, and that a 3D geometric modeling metaphor might better support human cognitive processes. Informal experiments in CAD modeling of sample geometric shapes reported in the sketch recognition literature shows that the two approaches are comparable. This evidence suggests that computer sketch recognition may be unnecessary, and that efforts should be directed toward improving the human factors aspects of current CAD software to better support the needs of conceptual design.

Modeling and Implementing Concurrent Engineering in a Virtual Collaborative Environment

Many companies are creating mterdisciplmary concurrent engineering teams to collaborate on product development projects from dispersed locations. The problems these companies face stem from the lack of understanding of technical, organizational, and pro cedural requirements for such activities to take place effectively. This paper proposes an integrated conceptual model of the product life cycle, and a virtual collaborative environment to support concurrent engineering. The product life cycle model features the life cycle pro cess itself, but also highlights the corresponding people and technology required for each phase. The virtual collaborative environment framework models the activities that occur, and the underlying infrastructure that must be present for collaboration to be effective. A sim ple prototype of the model is implemented, and an exploratory study is conducted to gain insight into the issues of concurrent engineer ing in a virtual collaborative environment, and to identify future research directions.

Analysis of the general 4R and 5R robots using a vector algebraic approach

Abstract The forward and inverse kinematic problems of the general 4R and 5R serial robots are analyzed using a vector algebraic method. The displacement equations of the general 4R and 5R robots are obtained as first and eighth order polynomials, respectively. The approach in this paper is shown to be simpler than other related approaches.

Kinematic analysis of certain spatial mechanisms containing higher pairs

Abstract Certain spatial mechanisms containing higher pairs are kinematically analyzed in this paper using the vector algebraic method . This method uses standardized vector expressions and operations, and specification of the relative orientation of the elements of higher pairs is not required. As a result, the analysis procedure is simpler and more direct than with other methods.

CMM sequence optimisation with collision detection

This paper presents a method for finding optimal collision-free inspection sequences for Coordinate Measuring Machines (CMMs). The sequencing problem is formulated as a standard Travelling Salesperson Problem (TSP). During the network construction, collision detection is performed for each pair of measurement points using a novel image-based collision detection method. Penalties are added to path segments with collisions, and the nearest-neighbour TSP algorithm is applied. For the path segments with collisions, a heuristic algorithm is employed to make a detour around the interference volume. The proposed methods are implemented using the ACIS solid modelling kernel and OpenGL graphics library. The effectiveness of these methods is verified by simulations to demonstrate the collision-free path generation of parts with complex geometry and a comparison of the TSP solutions with and without the collision penalties is presented.

Neural networks and the inverse kinematics problem

Inverse kinematics is a fundamental problem in robotics. Past solutions for this problem have been realized through the use of various algebraic or algorithmic procedures. In this paper the use of feedforward neural networks to solve the inverse kinematics problem is examined for three different cases. A closed kinematic linkage is used for mapping input joint angles to output joint angles. A three-degree-of-freedom manipulator in 3D space is used to test mappings from both cartesian and spherical coordinates to manipulator joint coordinates. A majority of the results have average errors which fall below 1% of the robot workspace. The accuracy indicates that neural networks are an alternate method for performing the inverse kinematics estimation, thus introducing the fault-tolerant and high-speed advantages of neural networks to the inverse kinematics problem.This paper also shows the use of a new technique which reduces neural network mapping errors with the use of error compensation networks. The results of the work are put in perspective with a survey of current applications of neural networks in robotics.

Learning to control dynamic systems with automatic quantization

Learning to control dynamic systems with unknown models is a challenging research problem. However, most previous work that learns qualitative control rules does not construct qualitative states; a proper partition of continuous- state variables has to be designed by human users and given to the learning programs. We design a new learning method that learns appropriate qualitative state representation and the control rules simultaneously. Our method can aggressively partition the continuous-state variables into finer, discrete ranges until control rules based on these ranges are learned. As a case study, we apply our method to the benchmark control problem of cart-pole balancing (also known as the inverted pendulum). Experimental results show that our method not only derives different partitions for the cart-pole systems with different parameters but also learns to control the systems for an extended period of time from random initial positions.