Bruce Kramer

Analysis of machine degradation using a neural network based pattern discrimination model

Abstract Machines degrade as a result of aging and wear, decreasing performance reliability and increasing the potential for faults and failures. The effect of machine faults and failures on factory productivity is an important concern for manufacturing industries. Economic effects relating to machine availability and reliability, as well as corrective (reactive) maintenance costs, have prompted facilities and factories to improve maintenance techniques and operations to monitor machine degradation and detect faults. This paper presents an innovative methodology to change maintenance practice from breakdown reaction to breakdown prevention, thereby reducing maintenance costs and improving productivity. To analyze machine behavior quantitatively, a pattern discrimination model (PDM) based on a cerebellar model articulation controller (CMAC) neural network was developed. A PUMA 560 robot was used to study the feasibility of the developed technique. Experimental results have shown that the developed technique can analyze machine degradation quantitatively. This methodology can help operators set up machines for a given criterion, determine whether the machine is running correctly, and predict problems before they occur. As a result, maintenance hours can be used more effectively and productively.

Development of tele-operated micro-handling/machining system based on information transformation

Describes the construction of a tele-operated handling/machining system which operates in the micrometer to millimeter range. The system is equipped with a pair of three dimensional force sensors and can be controlled remotely using a three dimensional joystick. The paper presents a transformation method for converting multi-axis force information to auditory information. Such a transformation conveys a realistic sense of the state of the operation and allows high operability of the system. Experimental results are presented which demonstrate the effectiveness of the implemented system.

A tele-micro-surgery system across the Internet with a fixed viewpoint/operation-point

This paper describes the necessary functions and the construction of a micro-surgery system which can be operated from a separate place. In particular, a vision system which has a fixed visual point and a slave manipulator, which has the same fixed point at the end of the tool, are proposed. In the vision system, the axes of all rotational degrees of freedom intersect at the focal point of the microscope. In the slave manipulator, the axes of all rotational degrees of freedom also intersect at this point. These mechanisms provide high accessibility to small objects independent of the approach angle from the microscope and the manipulator to the object. Another feature of the system is the transformation of multi-axis force to auditory information to enhance the operability of the system by indicating the contact state of the end-effector and object. Furthermore, the visual information acquisition and display systems are controlled in accordance with the inferred intention of the operator. Finally, a tele-micro-handling experiment was performed using the Internet and two artificial satellites.

Dynamics modeling of robotic manipulators using an artificial neural network

Dynamics modeling is important for the design, analysis, simulation, and control of robotic and other computer-controlled mechanical systems. The complete dynamic modeling of such systems involves the computationally intensive solution of a set of non-linear, coupled differential equations. Artificial neural networks are well suited for this application due to their ability to represent complex functions and, potentially, to operate in real time. The application of an artificial neural network to dynamics modeling of robotic systems is investigated. The Cerebellar Model Arithmetic Computer (CMAC) is employed. A hybrid implementation of CMAC is proposed to allow use of the model for either simulation or control of robotic manipulators. The success of the simulated results and the accuracy of the generated outputs after a few training cycles demonstrate great promise for further development of the method and its implementation in control systems.

Development of an inter-world tele-micro-surgery system with operational environment information transmission capability

This paper describes the necessary functions and the basic structure of a tele-micro-surgery system. In particular, the system is comprised of a vision system which is focused at the end of a surgical tool and a slave manipulator, the rotational axes of which intersect at the same focal point. These mechanisms realize high positionability that is independent of the approach angle from the microscope and the manipulator to the small object that is being operated on. The system also incorporates the transformation of multi-axis force information to sounds that can be easily interpreted by the operator to indicate the contact state between the tool and the object. Furthermore, an inter-world tele-micro-handling experiment has been executed successfully from a remote site using the Internet and two artificial satellites.

Experiments in tele-handling and tele-machining at the macro and micro scales, using the Internet for operational environment transmission

This paper describes the construction of a tele-handling and tele-machining system at the macro and micro scales that uses the Internet for communication. The necessity of transmitting information concerning the operational environment requires an information compression method based on the real-time judgment of the machining state, the transmission of the compressed information and the transformation and augmentation of the information for presentation to the operator. A method to compensate for transmission time delays using the physical model of the object is also described. Results of tele-handling and tele-machining experiments connecting Washington DC and the University of Tokyo in Japan are discussed.

A Manufacturing System for the Global Age

Abstract The concept and construction of an “inter-world manufacturing system,” which provides an efficient and convenient environment for designing, developing and manufacturing new products or research-related equipment, is described. The essential technologies which are needed to realize the inter-world manufacturing system are a human-friendly manufacturing system, including a multi-sensor integrated machining center, a capability for remote reality communication and a computer-supported cooperative work (CSCW) technique. In our Washington, DC - Tokyo telemachining experiment, a multi-sensor integrated machining center was successfully operated using these technologies. The experimental results of this test are described.

Development of a system for robotic deburring

Abstract Relatively crude robotic deburring has been accomplished through the detailed programming of the edge contour and the provision of significant passive compliance to compensate for positional errors. In precision deburring, the uncertainties in edge position relative to the tool location may be too great to allow a sufficiently uniform chamfer to be obtained using passive compliance alone. In addition, the costs associated with the detailed path programming can represent a significant fraction of the cost of deburring. This paper discusses the development of a robotic deburring system which provides for simultaneous edge following and force control. An AdeptOne robotic manipulator has been interfaced with an instrumented remote center of compliance manipulator to provide a system which tracks an edge at a controlled force level which can be specified by the operator.

A tele-machining system using multi-axis force data and stereo sound information

This paper describes a tele-machining system using multi-axis force data and stereo sound information. In the proposed system, the force information from the multi-axis force sensor is used not only to determine the machining and the operation state but is also reflected back to the operator through a three degree-of-freedom joystick to give the human operator the impression that he is directly manipulating the object. This illusion is created by mapping the multi-axis force information obtained in the cutting zone to the joystick system. As a further aid to understanding the process state, stereo sound, including both verbal and non-verbal information, is included in the system to enhance the user interface. Changes in the cutting state are perceived by the operator as changes in the tone quality. Verbal information is transmitted to the system by the operator in the case of an extraordinary cutting state and is also used to confirm the instructions of the operator. The availability of such verbal and non-verbal information decreases the anxiety of the operator. The effectiveness of the system was demonstrated by experiment.<<ETX>>

Intelligent materials processing

Abstract Current applications of knowledge-based expert systems in manufacturing involve decision-making that is based on information that is readily interpretable. In process control, the relationship between the underlying process behavior and the sensed information is not usually well understood. In addition, the required response speed of the system can be quite short; from a few milliseconds to a few seconds. The dual requirements for fast decisions and the incorporation of “fuzzy” knowledge requires new approaches to the development of expert systems for process control. Both advanced, fast executing artificial intelligence algorithms and improved, fundamental physical understanding of the underlying manufacturing processes that are being controlled are necessary. It is important to emphasize that neither the artificial intelligence nor the manufacturing engineering communities currently have the tools and techniques that are required to solve the full range of process control problems. It is essential that communication and cooperation between the two disciplines be increased to give a full appreciation for the current strengths and limitations of the respective science bases and to define a clear direction for advanced, fundamental research within each discipline.