This paper presents a new formulation as well as numerical solution for the problem of finding a point-to-point trajectory with maximum load carrying capacities for flexible manipulators. For rigid manipulators, the major limiting factor in determining the maximum allowable load (mass and mass moment of inertia) is the joint actuator capacity, while the flexibility exhibited by light weight robots or by robots operating at a higher speed dictates the need for an additional constraint to be imposed for situations where precision tracking is required, that is, the allowable deformation at the end effector. The Lagrangian assumed mode method was used to model the manipulator and load dynamics, including both joint and deflection motions. An Iterative Linear Programming (ILP) method is then used to determine the maximum allowable load of elastic robot subject to both constraints, while a general computational procedure for the multiple-link case given arbitrary trajectories is presented in detail. Symbolic derivation and simulation by using a PC-based symbolic language MATHEMATICA® was carried out for a two-link planer robot and the results further confirm the necessity of the dual constraints.

Rough joint flexibility is the dominant source of compliance in today's commercial robots in future robots containing light weight flexible arms link flexibility may become most important. Hence this paper stresses link flexibility rather than joint flexibility.

This paper presents a qualitative theoretical formulation for synthesis and analysis of multiple contact dexterous manipulation of an object, using a robot hand. The motivation for a qualitative theory is to build a formalisation of ‘human-like’ common-sense reasoning in robotic manipulation. Using this formalisation, a robot hand can perform finger-tip manipulative movements by analysing the physical laws that govern the robot hand, the object, and their interaction. Traditionally, such analysis have been framed in quantitative terms leading to mathematical systems which become intractable very quickly. Also, quantitative synthesis and analysis, often demand an accurate specification of the parameters in the universe of discourse, which is almost impossible to provide. The qualitative approach inherently encounters both these problems successfully.

The qualitative theory is presented in three developmental stages. A qualitative framework of spatial information in the context of dexterous manipulation has been provided. Qualitative models of an object configuration and transformations in them that occur during a manipulation process, have been developed. Finally, the development of a ‘quasi-static’ qualitative framework of a dexterous manipulation process that performs the desired object transformation, has been presented.

A method is presented to perform path planning in the Cartesian space. The motion velocity and acceleration are continuous. This is the first polynomial path planning algorithm in the Cartesian space working in real-time; it does not need an advanced knowledge of the trajectory and can be used when knot points are provided on-line by a sensor. The Bézier representation is used to compute a fifth degree polynomial path. It has been tested on an industrial robot controller.

Much of the work of Artificial Intelligence concerns the representation and symbolic manipulation of various domains. One problem shared by existing systems is their limitation to discrete states and state changes, brought about by the linking of declarative states with points in time. With this identification, models of continuous processes must forego such declarative states, resulting in a loss of decomposability that limits the amount of complexity that can be handled by a given amount of computation.

In this paper we develop a model for robot problem solving over a domain of continuous actions. The proposed world model consists of a vector of piecewise defined functions of time, i.e. functions that can be represented using finite sets of expressions and finite partitions of time intervals. The critical points of these functions are represented symbolically as discrete events. The goal is represented as a point through which the world model vector must thread at some point in model time. World model functions are transformed by adding events at various discrete points in time which transform the trajectories in the direction of the goal. As events are added, formulas relating the event times are added to the description of the world. The result is a partial order of events which represents a plan for “forcing” the world function to achieve the goal.

The relationships between this model and the classical problem-solving systems (e.g. STRIPS) are utilized in order to produce a planning procedure. This procedure is applied to a robot problem which can be solved using the model. We conclude by presenting an analysis and prospectus of considerations for an actual implementation of these ideas.

The concepts of entity-relationship diagram have been applied to picture description. Primitive picture entities, picture relationships, and picture grammars are presented with illustrative examples. A high-level description of a two-level picture generation system is proposed using either string description or ER diagram description. Illustrative examples are also given. The advantages of ER diagram description together with its comparison to string description are also presented. The results may have useful applications in robotics, artificial intelligence, expert systems, picture processing, pattern recognition, knowledge engineering and pictorial database design.

A dynamic model for two three-link cooperating structurally-flexible robotic manipulators is presented in this study. The equations of motion are derived using the extended Hamilton's principle and Galerkin's method, and must satisfy certain geometric constraints due to the closed chain formed by the two manipulators and the object. The dynamic model presented here is for the purpose of designing controllers. Therefore, a low-order model which captures all the major effects is of interest. Computer simulated results are presented for the case of moving an object along an elliptical path using the two cooperating flexible manipulators.

Local torque minimization for redundant manipulators has been considered by using a redundancy resolution approach. Since its initial formulation, however, proposed solutions have been plagued by performance instabilities. Till now, the reason was not clear. In this paper it is shown that the instability problem occurred because of incorrect formulation. A correct formulation is proposed, which is used to analyze the reason for instability. Special attention is paid to joint acceleration terms at the selfmotion manifold. This helps explaining the recent formulation of a stable scheme based on local joint velocity minimization in terms of torque. The results obtained. are related to dynamic redundancy resolution in general.

It is standard now in undergraduate and graduate courses in robotics to teach the basic concepts of position control design strategies. Due to the geared motors inherent in most educational and industrial manipulators, sophisticated control design strategies such as the inverse dynamics technique cannot be easily demonstrated in a laboratory setting. A direct drive 5-bar-linkage manipulator with reduced motor torque requirements is proposed in this paper for such a purpose. The manipulator dynamics are easily understood by undergraduates and an inverse dynamics control strategy is suggested which can be easily designed by students at the undergraduate level.

This note describes a new class of devices, corrugated polyvinylidene fluoride bimorphs. It describes the fabrication, testing and analysis of one simple configuration and discusses the general potential of the devices. As actuators, they offer promise for applications involving relatively large amplitude displacements and small forces with ease of control. As tactile sensors, the devices offer considerable promise due to their potentially high resolution, selectable compliance, sensitivity and durability.