This paper presents work done to control robots of different geometries and capabilities to complete a task neither is capable of independently. The algorithms were tested with a ‘cherry picker’ experiment that required one large, inaccurate robot to lift and carry a smaller, more accurate mobile robot in order to complete an inspection task by placing its end-effector at a specified distance from a visually-specified target. Both mobile robots are controlled using an uncalibrated visual guidance method. An overview of the algorithm for coordinating the two mobile robots is presented, along with details and results of experiments conducted to measure the accuracy with which the shared task is completed. Results have shown the system to be accurate and robust while requiring very little communication between the two robots.

This paper presents a vision-guided control method called mobile camera-space manipulation (MCSM) that enables a robotic forklift vehicle to engage pallets based on a pallet's actual current location by using feedback from vision sensors that are part of the robotic forklift. MCSM Is capable of high precision mobile manipulation control without relying on strict camera calibration. The paper contains development of the method as well as experimental results with a forklift prototype in actual pallet engagement tasks. The technology could be added to AGV (automatically guided vehicle) systems enabling them to engage arbitrarily located pallets. It also could be added to standard forklifts as an operator assist capability.

In this paper, we present a high-precision visual control method for mobile manipulators called mobile camera-space manipulation (MCSM). Development of MCSM was inspired by the unique challenges presented in conducting unmanned planetary exploration using rovers. In order to increase the efficacy of such missions, the amount of human interaction must be minimized due to the large time delay and high cost of transmissions between Earth and other planets. Using MCSM, the rover can maneuver itself into position, engage a target rock, and perform any of a variety of manipulation tasks all with one round-trip transmission of instruction. MCSM also achieves a high level of precision in positioning the onboard manipulator relative to its target. Experimental results are presented in which a rover positions a tool mounted in its manipulator to within 1 mm of the desired target feature on a rock. MCSM makes efficient use of all of the system's degrees of freedom (DOF), which reduces the required number of actuators for the manipulator. This reduction in manipulator DOFs decreases overall system weight, power consumption, and complexity while increasing reliability. MCSM does not rely on a calibrated camera system. Its excellent positioning precision is robust to model errors and uncertainties in measurements, a great strength for systems operating in harsh environments.

A third-order set of nonlinear, ordinary differential equations models the relationship between internally measurablewheel rotationsand the position and orientation of an automatically guided vehicle, buttheserelationships areimprecise, growing increasingly inadequateastheirintegrals, and thevehicle,proceedfrom pointofdeparture. An extended Kalman e lter (EKF) is used to combine video observations of features on that portion of the environment that does not move, together with the sensed wheel rotations, to produce the ongoing estimates needed for navigation. The experimental usefulness is examined of a byproduct of the e lter, the estimate error covariance matrix, to an integrally related process: the process of identifying video observations with features of known location within the environment; these identities are required for application of new vision observations to the stateestimates. Thegoodnessof theEKF’ sprobability density functions isexperimentally examined by comparing them against actual, accumulated data; experimental results are presented from the use of an extensive theoretical developmentthatassesses, basedonrelativeprobabilitiesinferredfrom thesedistributions, theidentitiesofdensely occurring, nondistinct cues.

An automatically-guided wheelchair system has been developed at the University of Notre Dame. Like some other automatically-guided vehicles (AGVs), this system uses filtering to combine information from measurements and the state equations to produce accurate ongoing pose estimates. This approach relies on the detection of nonunique passive visual cues placed at known positions throughout the environment. In order to maintain accurate pose estimates, a measurement involving any given detected cue must correspond with the correct cue identity. This paper deals with the approach taken to address the problem of establishing correct identity.

This paper describes the development of an automatically guided powered wheelchair for individuals with severe disabilities. The navigation and control of the wheelchair is based the accurate estimation of the location of the wheelchair within its operating workspace. A novel method used to generate and track reference paths which take the user to and from various destinations within the wheelchair's environment is presented. The paper also provides a qualitative description of the restrictions and requirements that are specific to the wheelchair application as well as the way in which the current system addresses these restrictions and requirements. Finally, actual experimental runs of the wheelchair system are presented.

The paper presents a robust and precise means of achieving ''semiautonomous'', vision-based robot control which may be applicable to a range of construction objectives. The method entails the designation of ''camera-space objectives'' for use with the method of camera-space manipulation, by an intermediate step involving the direction of a narrow beam of light. The method is shown experimentally to be effective without the need to calibrate the geometry of participating elements, including manipulator, cameras, or laser-pointer-bearing pan/tilt unit.

This paper presents an overview of the methods used with on experimental prototype of on autonomous powered wheelchair. The paper focuses on the importance of describing reference paths as continuous (or piecewise-continuous) geometric entities, rather than as a function of time or a series of discrete points. Furthermore, the importance of using the current estimated position and orientation of the wheelchair in order to select the reference point on the path is examined.<>