In robotic assembly operations the robot comes in contact with the environment and some form of compliance is typically used to prevent excessive contact forces. These contact forces provide information about the contact geometry and can be used to guide the assembly operation. Thus, compliance can be thought of as a task-dependent mapping from sensed forces to corrective motions which bring the robot closer to its goal. In this work we describe a methodology for the automated synthesis and verification of this mapping. The possible erroneous configurations of the robot are computed via Monte-Carlo simulation of the task execution. The static contact force at each erroneous contact configuration is computed and a corrective motion is computed and mapped to the force. The validity of the mapping is also checked to ensure that the sensed forces will provide unambiguous information during the actual execution of the planned compliant motion. Simulation and experimental results are also presented.

There are numerous applications in which robots are needed to assist rather than replace human operators. These applications typically involve either work in a dangerous environment or performing tasks that are physically impossible for humans to accomplish by themselves. Robots used in this mode do not need to be autonomous because they can essentially be controlled by the operator. Nonetheless, it is difficult for an operator to flawlessly control a robot with a large number of degrees of freedom. Therefore, an intelligent control system is needed to correct and improve the operator's input to the robot. In this paper the authors discuss a central component in this control system known as the variably-autonomous trajectory generator (VATG). The VATG combines input from the operator with input from one or more autonomous control modules (CM) to generate the robot trajectory. The amount of impact the autonomous CMs will have on the overall trajectory will depend on a number of factors such as the task being performed and the sensing modalities available, hence the authors refer to this type of control as being variably-autonomous.>

One factor that limits the range of tasks that robots can perform robustly is the scarcity of useful sensors available to provide feedback to the robot control system. While much progress has been made with vision sensors and force/torque sensors, tactile sensing systems have fallen behind, and so general-purpose tactile sensing systems are not commercially available. The aim of this work is therefore to produce a tactile sensing system that could be manufactured inexpensively, could be used on a wide variety of robotic systems, and would provide the kind of output typically required in dexterous manipulation applications. Such a tactile sensing system will be presented in this paper. The tactile transducers used in this system are based on semiconductive ink technology that allows transducers of any size, shape, and resistance range to be produced merely by altering the ink printing process and substrate geometry. Each sensor outputs three pieces of information that are useful in robotic manipulation-two parameters indicating the location of a contact point on the transducer and one parameter specifying the amount of force being exerted at the contact point. So that the sensing system could support transducers of different shapes and sizes, the analog interface circuitry of this system has been designed to be fully programmable, and also includes circuitry to enable self-calibration with appropriate software.>

In this paper we will discuss the control of a five-fingered, 18 degree of freedom, tendon driven robot hand known as the Atlas Anthrobot. Developing a traditional PID controller for this hand has proved difficult because the behavior of the tendon drive system produces is highly nonlinear. This nonlinear nature of the system however does suggest it as a candidate for fuzzy control. Along these lines we have developed a PID control system and a fuzzy control system for the hand which are both capable of simultaneously controlling the position of the fingertips and the amount of force exerted through a contact point on the fingertip. In this paper we will present a comparison between the two systems in order to generally make statements about the appropriateness of each controller for controlling complicated tendon driven devices such as our robot hand.>

The authors discuss a system, called AMP-CAD, which develops assembly motion plans using CAD models of parts. They place an emphasis on the calculation of the potential fields used for motion planning. Potential fields can be thought of as a mechanism to interface CAD representations with plan development. AMP-CAD is a complete system that combines an assembly representation, a two-phase motion planner, and a execution unit into a single integrated package. Implementation and experimental results are discussed.>

In traditional geometric modeling, it is common to assume that models of objects created are complete and correct. Unfortunately in real life robots and other autonomous machines may have only uncer- tain knowledge of only a portion of an object's geom- etry and therefore we need to be able to create object models that are incomplete and uncertain. Towards this end we have developed mechanisms for represent- ing geometric entities in terms of Dempster-Shafer belief functions. To do this we have extended the no- tion of belief functions to include continuous frames of discernment. A fractal model of the belief func- tions is developed and information theory concepts are applied for reasoning about a sensor data model vis-a-vis reference object models. An implementation of these concepts and experiments involving them are discussed.

A novel approach to inspection is pre- sented based on comparing a volumetric representa- tion of a reference object to a volumetric representa- tion of an actual object iteratively created from sensor data. The use of volumetric representations gives this approach a number of distinct advantages over the more traditional surface-based inspection methods. One significant advantage of volumetric inspection is that there is no need to identify geometric features, particularly important when inspecting objects such as castings and forgings whose features are often dif- ficult to identify. Secondly, the volumetric represen- tation lends itself naturally to a multi-sensor environ- ment. By this we mean that a volumetric model of the sensed object can be synthesized utilizing information from multiple sensors of varying types, accuracy, loca- tion, etc. This allows a choice of the most appropriate sensor for acquiring information about the different regions of the object. Finally, the volumetric represen- tation allows us to carry out true three-dimensional inspection as opposed to the typical approach where renderings of the reference model are compared to two-dimensional images of the actual object.

The authors introduce a planning system that relies on a novel approach which converts geometrical descriptions of assembly parts into potential field representations and then translates assembly operations into equivalent operations on the potential field representations. As a consequence, the system is able to quickly compute a path for the assembly motions and isolate those regions of the path where force/torque guided motions must be introduced to cope with possible collisions that may be caused by the prevailing uncertainties in the positions/orientations of the parts and their dimensions. This system has been experimentally verified using a Cincinnati Milacron T/sup 3-/726 robot to assemble a gear box.>

Vision sensor output can be processed to yield a multitude of low-level measurements, where each is inherently uncertain, which must somehow be combined to verify the locations of an object. It is shown that this combination can be accomplished via Dempster-Shafer theory using binary frames of discernment (BFODs). A special advantage of BFODs is the computational ease with which they allow information from disparate sources to be combined, which is particularly significant in light of recent concerns about the exponential complexity of a brute-force implementation of this theory.>

A method for the mating of tightly fitting parts in the presence of significant sensing, model, and control uncertainties is presented. Using this method randomly placed parts that were located using a three-dimensional laser range scanning system can be mated even when the clearance between them is only 0.001 in. The various sources of error introduced in the fine forced-guided motions used to execute assemblies are examined. Among these sources are noisy force/torque readings, mechanical vibrations, the presence of sliding and sticking frictions, and the possibility of eccentric oblique impacts. It is pointed out how most of these errors might be reduced or eliminated. Straight-line motion goals (SLMGs) are then processed by the basic building blocks of a dynamic planning strategy. By dynamically building an assembly plan out of SLMGs, it is possible to carry out robotic assemblies in the presence of sensing and model uncertainties while recognizing and recovering from errors introduced by control uncertainties. A specific instance of this approach is outlined.>

The authors present seldom-discussed sources of error in force-guided motions to execute assemblies. Among these sources are noisy F/T (force/torque) readings, the presence of sliding and sticking frictions, and the possibilities of eccentric oblique impacts. They point out how most of these errors might be reduced or eliminated. The authors introduce the notion of straight line motion goals (SLMGs) and show how SLMGs can be considered as the basic building blocks of a dynamic planning strategy. The steps required to carry out peg-in-hole experiments are examined. It is shown how the torque vector can be used to compute the direction in which the peg should be compliantly moved to align the axes of the peg and the hole. The authors present an error-detection-and-recovery algorithm for these experiments and show the dependence of the success rate on the number of error-detection-and-recovery cycles allowed.>