Robotics最新文献

The wabot-2 is an anthropomorphic robot playing keyboard instruments, developed by the study group of Waseda University's Science and Engineering Department.

The wabot-2 is equipped with hands tapping softly on keys, with legs handling bass keys and expression pedal, with eyes reading a score, and with a mouth and ears to converse with humans. Based on wabot-2, wasubot has been developed by Sumitomo Electric Industries Ltd., whose artistic skill has been demonstrated in performing music at the Japanese Government Pavillion in Expo'85.

The present paper summarizes the wabot-2's motion, visual and vocal subsystems as well as its supervisory system and singing voice-tracking subsystem.

An increasingly important topic in capital equipment acquisition is the analysis of robots in the manufacturing environment. Due to the unique characteristics of robots, it is important to recognize all the costs and benefits of a robotic installation in evaluating the feasibility of including this new piece of equipment in the production process. What makes analyzing the proposed robotic acquisition different from the usual capital equipment purchase is dealing with the many secondary and tertiary benefits and costs afforded by robotic systems.

The chief difficulty is estimating the costs and benefits due to the flexibility of the robot installation.

There are three important and essential steps in the determination of the value of a robotics installation. Initially it is necessary to make an informed estimate of the total investment required to obtain and install the equipment. Secondly, the effect of the incremental investment on the operation of the firm both in terms of expenses and profitability must be carefully measured. And finally, the third step is an analysis of the value of the equipment in terms of the return provided in relation to the required investment.

Task level programming systems previously proposed by several authors are briefly discussed and another approach is also offered.

The possibilities of integration of components such as vision and tactile modules, cad systems, robot simulators, specialized planners, etc., are analysed. This should help us to understand their interrelations and limitations when their integration in a system is attempted. The notion of robot station, rather than a mere manipulator, will be kept in view.

The conceptual structure presented has two phases:

• •

  - off-line: automatic plan generation (implicit programming) and plan testing (in simulated execution)

• •

  - on-line: execution supervision with sensorial feedback and local planning capability.

The usual limitations of an off-line approach are analyzed and a more active role of simulation of the whole station is suggested, specially in what concerns a comprehensive sensorial simulation, allowing moving to this phase most verifications, previously mandatory on the on-line phase. This should allow a more realistic plan generation and, eventually, interactive planning (with better debugging tools).

In the on-line stage, emphasis is placed on the relationship between execution supervision and sensorial feedback. One suggestion of automatic plan repair, valid for any of the phases is also presented.

Part manipulation by a robot is practicable provided the orientation of the presented part is known. Traditionally, a multi-part feeder will consist of either a number of dedicated feeders or else of plain feeders equipped with sophisticated electronics. The development of an inexpensive feeder with an uncomplicated recognizing device might be a good alternative. Two recognizing devices of this kind are described. One device includes a moving belt, 3 sensors, and a programmable logic controller, but the device is dedicated to one particular part. The other device consists of a moving belt, 16 sensors, and a microcomputer. This device has self-learning capability. Both devices are able to offer information about the orientation of the part that will be presented to a robot.

The continuously growing rate of production and use of industrial robots has proved that, in parallel with other implementation aspects, due attention should be paid to problems related to their operational safety. Safety has to be planned and built in at the level of robot workplaces.

Although safety problems are encountered by various groups of personnel, it is mainly the programmer and the trouble-shooter who are at risk.

There are lacunae in the safety provisions which could be bridged, on the one hand, by developing more appropriate safety equipment and, on the other, by imposing minimal safety requirements upon all robots as well as by giving examples for a safe design of robot workplaces.

A robot activity study was conducted for the U.S. Bureau of the Census in late 1985 to provide a detailed picture of the current and planned use of robotics in San Diego County, California. The results indicate an increasing usage rate, expectations of decreasing costs, and minimal employment effect to date, but perceptions of an increased impact. The current and future use of robotics in 65 companies is presented here as a series of charts.

Light-stripping, whereby a laser or other source of a stripe of light in conjunction with a single camera can provide three-dimensional information of a scene, can be used to estimate the volume of a seam gap for welding, filling, or gluing robots, and thus provide automated feedrate control. The geometry of light-striping systems is analyzed to yield an efficient method to map the locations of stripe pixels in the image to three-dimensional coordinates. Methods are discussed to determine fill-points by recognizing the pattern produced by seam walls despite noise. Combined with computation of areas and volumes, the steps simplify to an effective algorithm for microprocessor implementation.

An international conference entitled “Intelligent Autonomous Systems” was held in Amsterdam, The Netherlands from 8–11 December 1986.

Following the “Introductory Address” presented by chairman L.O. Hertzberger (University of Amsterdam, The Netherlands) there were eight invited papers which considered the theoretical and practical aspects of designing and operating autonomous systems. A further sixty-two papers were presented which provided an in-depth analysis of intelligent autonomous systems.

We present below a report on the lectures delivered at the conference.