ment are similar and relate to the efficiency and cost of task accomplishment. Typically, trade-offs consider factors such as man’s safety, equipment capital investment, personnel staffing, training, man versus machine reliability, maintenance and control system requirements. These factors are the same that we now face in the decision for a continued space presence and the move toward space industrialization. The scale of a future space construction and industrial project would require crews of astroworkers commuting routinely to job sites in Earth orbit. The cost of supporting such a large crew force and the logistics of the enterprise would be formidable. Precious time would be spent donning spacesuits, and once suited, useful astronaut energy is expended overcoming suit constraints. Also the astroworkers limited vision, strength, dexterity, endurance and consumable life support systems will almost certainly accelerate the need for mechanisation. As well, the foreseen construction in space produces a requirement for predictable, repetitive long duration tasks and such applications are most ideally accomplished by machines. Before future manipulator developments are discussed, it is perhaps timely that the capability and operational status of the existing space arm, the Shuttle Remote Manipulator System is reviewed. THE SHUTTLE RMS — CANADARM The arm is some 16 m long and 30 cm in diameter, and is primarily used to deploy and retrieve satellites and to support extra vehicular activity (EVA) servicing. Figure 1 shows the arm on-orbit manipulating a payload during the STS-3 mission. The arm represents the latest and perhaps the most significant of space manipulators. Two previous systems flown in space were task specific and controlled from Earth. Each collected soil samples, Surveyor 1967 from the moon and the Viking lander from Mars in 1976. The RMS, however, is a general purpose multidegree of freedom man/machine system, controlled by a complex set of nonlinear equations. Major features of the system are the arm control algorithms that allow a very smooth, easy to operate system even though the demands on the system can vary considerably. Variations occur because of the large range of payload mass properties and the many geometric arm configurations possible. With man-in-the-loop. Fig. 2, the system is designed for instinctive operation where the astronaut commands translational and rotational end-point rates using two three-degree-of-freedom hand controllers, which provide control without consciously controlling individual joints. Control algorithms convert input commands into resolved rate outputs for servojoints to produce end-point motions vectorially proportional to hand controller movements. The algorithms scale and limit the control inputs with the limits defined for the particular arm configuration and load. The end point of control can be selected with respect to either the orbiter coordinate system, end effector coordinates or a point within the payload (usually the centre of mass). The RMS software is configured to suit each payload. This ensures suitable arm response and control characteristics that are mission and/or payload specific. Endpoint position control can be implemented by recourse to the orbiter computer which can execute up to twenty trajectories. Alternatively, any set of coordinates within the reference system can be selected by keyboard entry into the orbiter computer which then drives the arm to the selected position. The payload, arm and orbiter’s
RkJQdWJsaXNoZXIy MTU5NjU0Mg==