LMF Mining Robots
5D.1 Mining Robot Functions
The requirements of seed mining robots which are components of the proposed growing, replicating Lunar Manufacturing Facility (LMF) include six basic functions:
Many additional mining-related functions conceivably could be performed by a robot system capable of the most general classes of excavation and mining activities. Indeed, such capacity might be absolutely essential if seed packages are dispatched to other planets than the Moon (e.g., Mars, Titan, Mercury, or Earth). These added functions include drilling, tunneling, blasting, and many others. But the basic six capabilities described above appear both necessary and sufficient for system survival and growth on the lunar surface.
5D.2 Design Alternatives
There exists a bewildering variety of mining and excavation machine technologies from which to draw in conceiving an autonomous vehicle (Nichols, 1976). The final design is a variant of the system devised by Carrier (1979) during a 1978 NASA-sponsored study on extraterrestrial materials processing and construction (Criswell, 1978).
In Carrier's system, strip mining proceeds in an annular sector ω radians wide as shown in figure 5.37. The total system is designed for gradual expansion. based on Earth supply or lunar colony supply, over a 30-year period. In the first few years of operation, all stripping and hauling to the central processing plant is performed by front-end loaders (also called, variously, the "shovel dozer.'' "dozer shovel," "tractor loader," "end loader," "front loader," or "loader"). These machines are used on Earth for digging, loading, rough grading, and limited hauling. In the lunar case, according to Carrier, the loader should be used at the outset for long hauling as the easiest way to start ore flowing into the central plant. After a few years the loaders may be augmented by a system of haulers, essentially large volume ore trucks carrying lunar topsoil back to the central plant. This permits the loaders to strip-mine full time.
Figure 5.37. -
Lunar surface strip mining.
Figure 5.37. -
While useful as a starting point in the present study, the Carrier system cannot perform all required LMF functions. Figure 5.38 shows the basic design for the LMF mining robot. This machine is a modified loader with a rollback bucket; has a dozer blade formed on the lower face of the loader bucket, reinforced so that the bucket can be placed in a locked, elevated position and the robot driven as a dozer; and has three attachments aft which are removed during normal work, including a precision grading blade with surface contour sensors, a simple tow bar, and a somewhat more versatile towing platform.
Figure 5.38. -
LMF mining robot design.
Figure 5.38. -
A loader equipped in this fashion should be able to perform all six basic LMF functions enumerated above. According to Nichols, in a pinch the mining robots should also be able to act as a primitive crane, as a more versatile variable blade pitch bulldozer, as a "reach down" dozer able to cut below the depth accessible to most dozers, and as a backdragger to smooth loose dirt. Finally, it should also be possible for two loaders to join face to face to lift large boulders which neither could conveniently lift alone.
5D.3 Mining Robot Design Specifics
The team considered various specific aspects of LMF mining robot design, including machine mass, power consumption, sensor configuration, and computational and information requirements. The results and conclusions are presented below.
Robot mass and power estimates. According to Carrier (1979), haulers may be much less massive on the Moon than on Earth since the lower gravity enables the same physical structure to carry more payload mass because the force per unit mass is less. In loaders, the vehicle mass is used as a counterbalance to prevent the machine from tipping over when fully loaded, so the mass relations for these machines change little from Earth in the lunar environment. Usual terrestrial practice is to multiply the bucket load mass by a factor of 2.0 to determine a safe tipping mass (the mass of the vehicle used as a counterweight). However, lunar equipment might incorporate automatic sensing systems to prevent tipping over so a safety factor of 1.2 should be sufficient (Carrier, 1979).
If the hauling mass per trip for all mining robots is Mh, m is the rate at which lunar materials must be mined to support the LMF replication schedule, and t is the time required for a robot to complete one cycle of operation (scoop up soil, deliver to LMF, return to pit), then Mh = mt. Using a factor of 1.2, the mass of mining robots is approximately M = 1.2 Mh = 1.2 mt.
Conservatively estimating an average of 40 km travel distance per round trip to the LMF per robot (from a 20 km radius annular pit surrounding the growing seed), an average transport speed of 10 km/hr, and a typical duty cycle of 50% for actual mining work (to leave time for repairs and nonmining labors such as grading, towing, or cellaring), then the mean cycle time
r = (40 km)(3600 sec/hr)/(50%)(10 km/hr) = 28,800 sec
The annual lunar soil hauling requirement is approximately 4X106 kg (see app. 5E) to replicate a new 100-ton seed each working year, so,
m = (4X106 kg)/(3.14X107 sec) = 0.127 kg/sec
Hence, mining robot mass is
M = (1.2)(0.127 kg/sec)(28,800 sec) = 4400 kg
(Approximately 4400 kg/1.2 = 3700 kg of lunar material are transported each cycle.) Note that M is the total mass of robots required, not necessarily the mass per robot. In fact, it is essential that the seed carry at least two such machines so that strip mining can proceed almost continuously given a 50% duty cycle and so that a "spare" is always available in emergency situations. Assuming linear downscaling the mass of each robot is 2200 kg.
In Carrier's strip-mining system the machines require an average of 0.3 W/kg. Mostly this is due to the hauling function, the most energy-intensive operation performed. Hence each mining robot requires about 660 W which may be drawn from 4 m2 of photovoltaic solar cell panels mounted on every available surface. A fuel cell module (Fickett, 1978) is included in the robot design, for buffer storage and peak load coverage when power consumption may rise as high as 10 kW (as during rescue operations). This module may be recharged at any time from the LMF power grid, but this should not be necessary as the robots should be fully self-sufficient in this regard. Finally, an electrostatic lunar dust wiper is provided to maintain solar cells and camera lenses at maximum efficiency.
Sensor configuration. Sensing equipment on board includes the usual navigational receiver which ties into the high-accuracy transponder network; a two-axis level sensor so the robot knows its tipping angle with respect to the local gravity field; a detachable grading sensor which rolls along the ground just in front of the precision grading blade and provides immediate real-time feedback to permit exact control of grading angle, pitch, and slew.
The most complex sensor system is the remote camera arm. (See discussion of state-of-the-art techniques by Agin, 1979.) The camera is binocular to allow ranging and depth perception, and to provide a spare in case one camera "eye" fails. This is mounted on a long robot arm which can be directed to observe any part of itself or to survey the landscape during roving activity. The camera arm will need at least seven degrees of freedom - rotation of the arm shaft, flexure of the two intermediate joints, bending at the wrist, camera rotation, lens rotation for focus, and telephoto capability for close scrutiny of interesting features in the environment.
The mining robot camera arm is absolutely essential if the vehicle is to function in the versatile manner envisioned for it. It is not enough simply to know position in space, because the environment in which the system must operate is highly complex. It might be possible for the seed computer to give the robot a "road map" to 1m accuracy, but this would not allow for proper navigation once the miners begin to physically alter their surroundings by digging, hauling, dozing, etc. Also, there may be objects smaller than 1 m that could cause major difficulties such as crevasses and boulders. Hence, it seems necessary to give the mining robots a true generalized "intelligent" roving capability.
Automation and AI requirements. The camera arm will require some high-level AI that lies beyond state-of-the-art. The onboard computer must keep track of the position of the moving arm in order to know where the camera is at all times. There must be routines for avoiding obstacles - for instance, the system should avoid hitting the camera with the loading bucket. Complex pattern recognition routines must be available to permit image focusing, telephoto operation, interpretation of shadows and shapes, differentiation between protrusions and depressions in the surface, and intelligent evaluation of potential risks and hazards of various pathways and courses of action. The onboard computer must have an accurate representation of its own structure stored in memory, so that the camera may quickly be directed to any desired location to inspect for damage, perform troubleshooting functions, or monitor tasks in progress. Finally, the computer must have diagnostic routines for the camera system, in case something simple goes wrong that can easily be corrected in situ without calling for outside assistance.
According to Carrier (1979) the automatic haulers can easily be designed to operate in an automatic mode, requiring only occasional reprogramming but substantially more advanced AI pattern recognition systems. (In 1980 a child's toy was marketed which can be programmed to follow simple paths (Ciarcia, 1981; "Toy Robots," 1980).) Carrier suggests that since there are so many variables associated with excavation "it is doubtful that the front-end loader could operate automatically," though the team disputes this conclusion. In addition to sophisticated pattern recognition and vision systems (Williams et al., 1979), the robot miners need a "bulldozer operator" expert system of the kind under development at SRI for other applications (Hart, 1975, and personal communication, 1980). Such an expert system would embody the knowledge and skills of a human excavator and could substitute for human control in most circumstances. In addition, expert systems might be executed remotely by a process called "autonomous teleoperation." In this mode of operation, mining robots can be remote-controlled via transponder network links by the master LMF computer, thus reducing onboard computer complexity.
Additionally, the onboard computer must handle such comparatively mundane chores as clocking, operating drive trains on the wheels, turning controls, blade angle control and configuration, task completion testing and verification, guidance and navigation, and internal diagnostics. An executive program is also required, capable of accepting new orders from the central LMF computer (e.g., "rescue machine X at position Y") and semiautonomously calculating how best to execute them (Sacerdoti, 1980).
Computation and information requirements. A first-cut estimate of the computational capacity required on board reveals that three major computer subsystems are involved: (1) robot camera arm (seven degrees of freedom, binocular vision, rangefinding, sophisticated AI such as pattern recognition and inference); (2) excavator expert system (controls physical operations, understands a world model, has expectations about outcomes, and can troubleshoot simple problems); and (3) high-level executive system (reprogrammability, interpretation, and "common sense" reasoning). Each of these subsystems represents a different problem and must be separately analyzed.
The robot system with mobile camera studied by Agin (1979) engaged in very primitive pattern recognition. This included insertion of bolts into holes, positioning a movable table relative to a fixed camera, velocity tracking (a Unimate PUMA arm, camera in hand, follows an object moving past on a conveyor belt), spot welding on a moving assembly line, and following a curved path in three dimensions at constant velocity (simulating industrial activities such as gluing, sealing, and seam-following). Again's visual recognition routines ran on a PDP-11/40 minicomputer, a 28K application, and the PUMA robot arm was controlled by the usual LSA-11 microcomputer which has a 16K capacity using 32-bit words. The visual system for the proposed mining robot will be at least 1-2 orders of magnitude more complicated than Agin's system, so we would estimate a control requirement of 106-107 bytes, or about 107-108 bits of computer capacity.
The SRI expert system "PROSPECTOR" runs on a DEC-10 computer with a 150K operating program and a 1M database, a total of about 3.2X107 bits (Hart, personal communication , 1980). PROSPECTOR "knows" about 1000 different factors related to prospecting. It is difficult to imagine a general excavation expert system requiring more than ten times this, or 10,000 factors, to achieve adequate autonomous operation with troubleshooting capability - the PROSPECTOR expert has generated some impressively accurate results in searches for ore-bearing bodies. If the "EXCAVATOR" expert system is thus about one order of magnitude larger than PROSPECTOR, the basic computational requirement is 10M or 3.2X108 bits.
Mining robot executive computer requirements are more difficult to estimate, as there are few previous directly applicable models. A simple passenger aircraft autopilot probably will run on a 32K microprocessor, and a "smart rover" vision-equipped wheeled mobile robot with a 6-degree-of-freedom arm developed in the 1970s at JPL used state-of-the-art microprocessors. Remarks by Sacerdoti (1979, 1980) on the subject of autonomous planning and execution in robotics suggest that the system required for robot miners is perhaps 1 to 2 orders of magnitude beyond current technology; thus the executive system may require a memory capacity of about 1 to 10M, or 3 to 30X107 bits.
Summing the requirements for the three major computer subsystems gives an "information bandwidth budget" of 3.6-7.2X108 bits, centering on about 500 Mb. The information necessary to completely describe the system for purposes of self-replication is probably on the order of 109 bits.
5D.4 LMF Approach and Access Geometry
In the baseline LMF scenario, mining robots must assume all hauling duties beyond the factory platform. Thus, it becomes necessary to specify how these mobile machines, normally bearing loads of strip-mined soil to be processed, will approach the factory and deposit their cargoes at an appropriate input location. A related query is how and where robots will accept waste products for transport to the pit for use as landfill. These questions are of some importance, because as the seed expands to full maturity it may become physically more difficult to exchange raw materials and wastes with interior LMF processing systems unless the access geometry has been designed to accommodate growth.
The solution adopted by the team is to earmark a constant-angle wedge corridor for permanent use as a mining robot access road. A 5° angle provides a corridor width of 5 m at the perimeter of the initial 60-m radius seed comfortably enough room for a mining robot to enter, drop off its cargo, pick up a load of waste materials, and then withdraw. The area of the constant-angle corridor increases as R2. This is the same dependence on radius exhibited by the area of the growing "seed," hence LMF mass, raw materials requirements, and waste production will increase at the same rate as the access corridor which supports interior factory systems. In other words, the expanding corridor prevents internal LMF systems from becoming "landlocked" as seed mass and radius grow exponentially in time.
The wedge corridor geometry is shown in figure 5.39. Note that as the LMF grows larger, mining robots (or any other external transport vehicle) must traverse ever greater distances, on average, to reach the entry corridor. For this reason a minimum of two such corridors should be provided, with the factory organized as two identical halves as suggested in figure 5.19. Further studies will be required to determine the optimum access and LMF configuration geometry from the standpoint of scheduling, efficiency, and access time.
figure 5.39. -
LMF constant-angle wedge corridor access route.
figure 5.39. -
Mining robots deliver raw materials to an input hopper located in the chemical processing sector, as shown in figure 5.40. Outshipments of waste materials are delivered to them in similar fashion. These hoppers serve as materials depots, able to help sustain LMF operations during periods when the supply of lunar topsoil is interrupted for any reason. Since each of the two initial seed robots makes one round trip about every eight hours, a hopper intended to serve as a one-week buffer must have a capacity of 42 mining robot loads or 76,900 kg of lunar regolith. A roughly cubical hopper constructed of 1 cm sheet aluminum and able to contain the weekly input volume of 42.7 m3 has a mass of 1650 kg.
Figure 5.40. -
Raw material delivery to input hopper.
Figure 5.40. -
Agin, Cerald J.: Real Time Control of a Robot with a Mobile Camera. SRI International Technical Note 179, February 1979.
Carrier, W. D., III: Excavation Costs for Lunar Materials. Fourth Princeton/AIAA Conference on Space Manufacturing Facilities, Princeton, New Jersey, 14 May 1979, AIAA Paper 79-1376.
Ciarcia, Steve: A Computer-Controlled Tank. Byte. vol. 6, February 1981, pp.44-66.
Criswell, David R.: Extraterrestrial Materials Processing and Construction. NASA CR-158870, 1978.
Fickett, Arnold P.: Fuel-Cell Power Plants. Scientific American, vol. 239, December 1978, pp. 70-76.
Hart, Peter E.: Progress on a Computer-Based Consultant. SRI International, Menlo Park, January 1975.
Nichols, Herbert L., Jr.: Moving the Earth: The Workbook of Excavation. 3rd ed. North Castle Books, Greenwich, Conn., 1976.
Sacerdoti, Earl D.: Problem Solving Tactics. SRI International Technical Note 189, July 1979.
Sacerdoti, Earl D.: Plan Generation and Execution for Robotics. SRI International Technical Note 209, 15 April 1980.
Toy Robots Gaining Intelligence. Byte, vol. 5, October 1980, p. 186. See also "Inside Big Trak," Robotics Age, vol. 2, Spring 1980, pp. 38-39.
Williams, D. S.; Wilf, J. M.; Cunningham, R. T.; and Eskenazi, R.: Robotic Vision. Astronautics and Aeronautics, vol. 17, May 1979, pp. 36-41.