TN295 No. 9079 40 • > » . 1 A~ O » „ ' • • ** .0^" o 'cr.T' " a, * ^ >«^J^° vv ^ , l -\ ^ -^ "»bv* ". "^^ V^ 1 r s ■-^8**" .^\ • i w ? ^ ^ v \ ' 0? ..-. °^ *°-° A°° *+ ^V .••• .^ ^^ - s°* O «"* , « o . *x, A v Xfc S * • , ^-v -A v*. ^d« <^ v ^t:t- ^0 S°* ^°^ < > . » • • ^ov* V /\ --^R : ***** %■ <5> * o * o * <$? O ►,„■ w A . t • . , **> c ,0 V \5 *^T* A <. 1 p v »b>* ^*v>*^ "w'^w'ii^^ v "^'"^ IC 9079 Bureau of Mines Information Circular/1986 Improved Backup Alarm Technology for Mobile Mining Equipment By Guy A. Johnson, Russel and Linneas W. Laage E. Griffin, (^ UNITED STATES DEPARTMENT OF THE INTERIOR Information Circular 9079 It Improved Backup Alarm Technology for Mobile Mining Equipment By Guy A. Johnson, Russell E. Griffin, and Linneas W. Laage UNITED STATES DEPARTMENT OF THE INTERIOR Donald Paul Hodel, Secretary BUREAU OF MINES Robert C. Horton, Director ins .114 Library of Congress Cataloging in Publication Data: Johnson, Guy A Improved backup alarm technology for mobile mining equipment. (Information circular ; 9079) Supt. of Docs, no.: I 28.27: 9079. 1. Mining machinery —Safety measures. I. Griffin, Russell E. II. Laage, Linneas W. III. Title. IV. Series: Information circular (United States. Bureau of Mines) ; 9079. TN295.U4 [TN3451 622s [622\8] 86-600017 CONTENTS Page Abstract 1 Notice 2 Introduction 3 General principles — design criteria 4 Proof-of-concept testing of prototype systems 5 Infrared systems tested 6 STI Omniprox 3070 6 Search-Eye 10 Ultrasonic sensing system 10 Doppler radar system 13 Possible hardware improvements 15 installation tips 17 Discussion 17 Infrared systems 17 Ultrasonic sensing systems 18 Doppler radar systems 18 Conclusion 18 Appendix. — Sources of equipment « 19 ILLUSTRATIONS 1 . Discriminating backup alarm design 4 2 . Ideal detection zone 5 3. Infrared sensor mounted on front-end loader 7 4. Infrared electronic control card 7 5. Infrared sensor on a Terex 90C 8 6. Close-up of infrared system sensor on a Terex 90C 8 7. Infrared sensor system on a Clark 275 9 8. Infrared system on a Caterpillar 980 9 9. Schematic diagram of an ultrasonic detection system 11 10. Ultrasonic detection system installed on a Clark 275C 12 11. Close-up of ultrasonic sensors 12 12. Measuring the detection zone 13 13. Close-up of a Doppler radar detection system 14 14. Detection zone of a Doppler radar system 15 15. Testing a Doppler radar system with an anthropomorphic dummy 16 16. Prototype synthesized voice warning system.... 16 TABLES 1. In-mine tests of discriminating backup alarm systems 6 2. Characteristics of backup alarm systems tested 6 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT ft foot rpm revolution per minute ft2 square foot st short ton g gram V volt GHz gigacycle per second V ac volt, alternating current h hour V dc volt, direct current m meter yd yard ms millisecond yd 3 cubic yard mW milliwatt yr year pet percent IMPROVED BACKUP ALARM TECHNOLOGY FOR MOBILE MINING EQUIPMENT By Guy A. Johnson, Russell E. Griff in, and Linneas W. Laage ABSTRACT Despite the use of warning alarms to alert miners to the backward movement of large mining equipment, miners still are injured too fre- quently in backup accidents. Currently approved backup alarm technology consists of continuous sounding alarm systems. New technology has de- veloped a warning system that initiates an alarm only if there is an object close behind the vehicle. This advancement eliminates constant exposure to the alarm because a warning is given only in case of a po- tential collision. Adoption of this development can improve safety and reduce damage, especially for front-end loaders (FEL's). It will also eliminate a source of nuisance noise in urban and residential areas. This Bureau of Mines report describes the general characteristics of infrared (IR) light, ultrasonic wave, and Doppler radar technology as used for backup alarms, and reviews the testing of the most promising detection devices. Supervisory mining engineer (now with Denver Research Center, Bureau of Mines, Denver, CO. ) . ^Electronics engineer. ^Mining engineer. Twin Cities Research Center, Bureau of Mines, Minneapolis, MN. NOTICE IMPROVED BACKUP ALARM TECHNOLOGY CAN BE APPLIED IN TWO WAYS 1. As a supplement to conventional backup alarms by sounding an additional alarm in the cab when a collision hazard is detected. 2. As a switch for the conventional backup alarm, sounding the alarm only when a collision hazard is present. With conventional backup alarms the oper- ator is expected to work the machine despite the alarm. With the improved backup alarm both the potential victim and the operator can respond to avert a collision because the false alarm aspect is eliminated. CAUTION The following statement was prepared by the Director, Office of Standards, Regulations, and Variances, Mine Safety and Health Administration, for instal- lations where the conventional backup alarm is switched on by the improved backup alarm (method 2 above). "Since the discriminating backup alarm does not give an automatic warning when the mine machine is put in reverse, it cannot be used to satisfy existing Federal requirements unless a petition for modification has been issued by the Mine Safety and Health Administration; therefore, anyone who would like to use the device will have to file a petition. This petition must be in writing to the Assistant Secretary of Labor for Mine Safety and Health." "The petition must contain the name and address of the petitioner; mailing address and identification number of the mine or mines affected; the mandatory safety standard to which the petition is directed and a concise statement of facts that would warrant the proposed modification." CAUTION These systems must be installed and aimed properly to detect collision hazards. These systems must be regularly inspected for operation and performance. INTRODUCTION The Mine Safety and Health Administra- tion (MSHA) (30 CFR 56.9087, 57.9087, and 77.410) requires "an automatic reverse signal alarm" on mobile surface mining equipment. These alarms usually are loud horns or bells on the rear of the equip- ment which are activated and in continu- ous operation when the vehicle's trans- mission is shifted into reverse. Their purpose is to warn miners of the rearward movement of the vehicle. As part of its research to develop col- lision protection technology for larger mobile mining equipment, the Bureau of Mines analyzed the occurrence of backup accidents. Bureau investigators reviewed fatality reports and held informal dis- cussions with such safety organizations as the Lake Superior Mines Safety Coun- cil, the Association of Arizona Mine Safety Engineers, the Wyoming Chapter of the American Society of Safety Engineers, and the Safety Section of the National Sand and Gravel Association. Results of this study show that back-over accidents still occur regularly, apparently because miners can become "immune" to the sound of current backup alarms . Semi continuous backup alarms inevitably produce the effect of a "false alarm," and repeated false alarms desensitize people exposed to them. The U.S. Army recognized this in Military Standard 1472: "The design of audio display de- vices and circuits shall preclude false alarms . "4 In mines, workers are constantly ex- posed to the repeated sounds of many backup alarms on various pieces of equip- ment. Consequently, they tend to pay little attention to any one specific warning. This is especially true for maintenance and ground crew personnel who constantly work near FEL's. To counteract this problem, the Bureau of Mines has adapted and developed sen- sors that will automatically turn on a backup alarm only when some object ^U.S. Army. Human Engineering Design Criteria for Military Systems, Equipment and Facilities. Military Standard 1472C, 1984, p. 51. (collision hazard) is behind a vehicle at a distance of 15 to 20 ft. If a warning is given only when the danger is real, it will get a greater response both from miners on the ground and from equipment operators. The Bureau reviewed the literature to determine the maximum proximity range at which sensors could be made rugged enough to be reliable in mines at a reasonable cost. Several options were found. Re- cent advancements in microcircuitry tech- nology make it easy to fabricate the electronic components for area and driver warnings once the proper detection sen- sors have been selected. Military-type detectors were studied first and found to be reliable but very expensive. Viability and low cost were both essential, and the fast-growing field of security and intrusion detectors offered promise. The security field has created a large market for sensors that give close-in object detection; for exam- ple, a sensor with the ability to detect the presence of a person sneaking across a room. Because of the economics of scale inherent in this market, relatively sophisticated sensing technology is now available at a very reasonable price. In the early 1980' s, the Bureau began testing, first in the laboratory and then in the field, prototype devices with a potential for solving the backup colli- sion problem. Although improvements in mirrors, the development of blind area viewers, and advancements in closed cir- cuit television already had made it easier for the operators of large equip- ment to see potential hazards, alarms still were needed to direct the attention of drivers, and potential victims, to specific dangers. In addition to solving the "false alarm" problem, an alarm that sounds only when a hazard exists can help to minimize the nuisance noise that extends to a mine's surrounding environment. Such noise is especially irritating in the early morning and evening hours around urban crushed stone or sand and gravel pits. Even though the miners often do not react to the backup alarms , people living close to the operations are constantly bothered by the noise. Alarms that sound only when something or someone is behind and close to a vehicle will be a boon to both mines and their neighbors. GENERAL PRINCIPLES— DESIGN CRITERIA In addition to the considerations pre- sented above, a few key principles have emerged from the Bureau's research into improved protection for backing vehicles. The "target area" behind the vehicle is relatively close, around 15 to 20 ft. (See figure 1.) Objects beyond this area are not a major hazard because they can be seen either in the vehicle's mirror or directly by the driver when the backing vehicle is turning. The rear-looking detector(s) are not just for sensing individuals or other vehicles. These systems can detect objects as small as 1 ft 3 (e.g., boulders, stacks of mainte- nance materials, etc.) and warn of their presence. This warning is accomplished when the hazard detection system initiates an in- cab backup horn or buzzer to alert the driver. (See figure 2.) An in-cab buzz- er can be added to the system when vehi- cle cabs are acoustically isolated from the external backup alarm. This in-cab warning feature uses modern, miniaturized electronic components that are low in cost. It is also cost effective because it alerts the driver to the possible presence of small objects that could dam- age the vehicle's tires, thus avoiding both the high material cost and lost time involved in replacement. During the Bureau's research, several alternative technologies were examined. In 1982 and 1983, after a few simple in- frared (IR) obstacle detection devices had been tested, a rugged unit produced by Scientific Technology, Inc. (STI), 5 Mountain View, CA, was found to give ade- quate coverage and distance during laboratory testing. (The IR sensor op- tion is discussed in the next section. ) -'Reference to specific not imply endorsement by Mines. products does the Bureau of -Transducer Backup alarm coverage area FIGURE 1.- Discriminating backup alarm design. I"-" FIGURE 2. - Ideal detection zone. Polaroid has packaged its ultrasonic ranging transducer (used on its cameras) for general use. This device was tested on a 24-yd 5 -capacity FEL and a 120- st-capacity haulage truck at the Bureau's Twin Cities (MN) facility, but the beam pattern was found to be too limited in diameter for on-vehicle use. (Followup work with a more sophisticated, long- range, ultrasonic system prototype for use on small construction equipment is also reported below.) The most recent work involves a newly developed short-range Doppler radar unit. In the late 1970' s, radar collision pro- tection was tested but found only practi- cal for long-range (100- to 300-ft) ap- plications because of signal processing circuitry limitations. The new Doppler radar systems are designed for close (4- ft) collision protection, so increasing the range is now a problem. Once the hardware modification problems can be worked out of this unit, it will be the most promising alternative yet tested. PROOF-OF-CONCEPT TESTING OF PROTOTYPE SYSTEMS The work detailed here is an extension of the Bureau's earlier efforts to use state-of-the-art technology in solving visibility problems inherent in large surface equipment. Table 1 summarizes the testing of discriminating backup alarms. Table 2 summarizes the charac- teristics of the backup alarm systems. TABLE 1. - In-mine tests of discriminating backup alarm systems System type and Testing Equipment System Comments location dates tested on range, 1 ft Infrared: Limestone quarry Mar. -Sept. Terex 90C. Up to 40. . . System had to be removed (MD). 1983. and made more rugged. Sand mine (MN) . . Oct. -Dec. 1983. Clark 275.. Up to 40. . . System removed owing to light sand-sun reflection problems. Sand and gravel July 1984- Caterpillar Up to 40. . . Marginal usage because of pit (CO). present. 980. moisture condensing inside lens. Ultrasonic: Limestone quarry Mar. 1984- Caterpillar Up to 17... Working well. (MD). June 1985. 988B. Sand mine (MN).. May 1984- present. Clark 275C. Up to 17.. . Working well. Doppler radar: Sand and gravel Mar. 1985- Caterpillar Up to 18... Initial interference be- pit (MO). present. 992. tween back-up alarm and cab annunciator. Correct- ed by a modified unit. Sand and gravel Feb. 1985 Caterpillar Up to 18. . . Switch problem at pit (CO). (1 day). 988. installation. Sand mine (MN) . . May 1985- present. Clark 275C. Up to 20. .. Maintenance problem, which was resolved. 1 Depending on size and reflectivity. TABLE 2. - Characteristics of backup alarm systems tested System type and manufacturer Shape of area coverage pattern Performance affected by — Infrared: Scientific Tech- nology, Inc. Ultrasonic: Global Fabrica- tions Co. , Ltd. Doppler radar: Con-Serv, Inc, 2 parallel narrow cones extending from sensors, 11- by 17-ft rectangle. 12- by 20-ft elongated teardrop. Sunlight, dust, reflectivity of object. Airflow, acoustical reflec- tivity of object. Radar profile of object. INFRARED SYSTEMS TESTED STI Omniprox 3070 The STI Omniprox 3070 series sensor was the first of the IR type to be laboratory and field tested. This sensor is a solid-state, modulated, IR beam detection and control device, provided in a modular configuration. The sensor head is total- ly sealed and shock tested (100 g at 10 ms). The manufacturer states it can be mounted anywhere — indoors or outdoors , submerged or in a vacuum — and can be lo- cated up to 30.5 m (100 ft) from the con- trol electronics. (Figure 3 shows the unit mounted on a FEL.) It is capable of disregarding ambient light (though direct sunlight does affect it) , atmospheric contamination, and thin film accumula- tions of oil, dust, water, and other air- borne deposits. Field tests support these latter claims. FIGURE 3. - Infrared sensor mounted on front-end loader. FIGURE 4. - Infrared electronic control card. The system is limited in its area of coverage. The 3070 series has an adjust- able range of up to 3.7 m (12 ft) in the proximity mode (target size 12 in) and 18 m (60 ft) in the retroref lective mode. The range sensitivity is adjustable through a potentiometer on the control electronics. The maximum IR beam is typ- ically 2-ft diam at a distance of 20 ft, which is too narrow for general in-mine use. A variety of standard output and con- trol options are available to adapt the system for different applications. For the Bureau tests, however, it was used with two sensors with overlapping beams , in a logic "or" mode with a double-pole, switched-relay output. With this setup, either sensor detecting an object can ac- tivate a backup alarm, an audio in-cab alarm, or a light, in any combination de- sired. It is also designed to easily adapt to operation on 115 or 230 V ac, 10.5 to 13 V dc, or 24 V ac or dc. For testing purposes, the electronics control circuit card was ordered unmounted. It was then adapted for 24-V dc operation, and mounted in a National Electrical Man- ufacturers Association (NEMA) 12, type D enclosure, with appropriate holes added for power and sensor leads (fig. 4). The system was field tested in three different locations over a period of sev- eral months. The first test was on a Terex 90C FEL. At the time, the 90C was a preproduction, pilot model machine, us- ing an 8- to 11-yd 3 bucket on a field trial and demonstration. The sensors were mounted on adjustable brackets weld- ed to the frame on either side of the radiator grille, about 6 ft above ground level. (See figures 5 and 6.) The con- trol box and annunciator were mounted in the cab on a dash panel located to the right of the operator. Connecting cables between the two, and a cable connecting the backup alarm to the control box, were run under the loader deck alongside FIGURE 5. - Infrared sensor on a Terex 90C. FIGURE 6. - Close-up of infrared system sensor on a Terex 90C. FIGURE 7. - Infrared sensor system on a Clark 275. FIGURE 8. - Infrared system on a Caterpillar 980. 10 existing cable and hose runs. The con- necting cables for the sensors, which consisted of two-pair shielded conductors (Belden 8723) , were supplied by the manu- facturer. For convenience of installa- tion, connectors were inserted in the sensor cables approximately 5 ft from the sensor. Power was obtained from a ter- minal on the oil pressure switch, which meant the system would be on if the ve- hicle's engine were running. The FEL's backup alarm was connected to the system to sound if the sensors detected an ob- ject in their coverage area, regardless of the direction of the FEL travel. The IR system was on the FEL for ap- proximately 6 months during the early spring and summer of 1984. Throughout the testing, the IR system revealed many weaknesses. Reports from the mine indi- cate that after 2 weeks of use, the sys- tem became sensitive to the vehicle's en- gine speed. Above 1,200 rpm it worked properly, but below this speed it had no sensitivity and false-alarmed. Dust ac- cumulation on sensor windows caused the sensors to register a loss of sensitiv- ity, but this condition was easily cor- rected by cleaning the windows. Despite these flaws, the principal operator of the FEL liked the system because it was useful in stockpiling operations and maneuvering near the highwall. The oper- ator also felt the system had prevented a couple of collisions. At the conclusion of the testing peri- od, an examination of the system revealed that the sensor leads between the control electronics and the sensors were worn and abraded in several places, causing short circuiting of the leads. This accounted for the poor operation of the system and indicated the need for better routing and securing of the leads. The Bureau has mounted similar STI sys- tems on smaller FEL's: a Clark 275 (fig. 7) and a Caterpillar 980 (fig. 8). The installations were much the same as pre- viously described except for the physical placements of the sensors and control electronics, which were located according to the situation and available cab space. Flexible, watertight conduit was used in the Caterpillar 980 installation to pro- tect the leads where they ran between the cab and the sensor location. These systems were affected by dust, sunlight, and reflected light from white sand, mak- ing their overall performance marginal. Sear ch-Eye Another IR based system, known as Search-Eye, manufactured by Global Fabri- cations Co., Ltd., in Weston, Ontario, Canada, was also tested. This system was designed primarily for use on street and alley refuse haulage trucks. It operates on the same principle as the previously described STI system. Reflected IR light from the detected object is sensed and activates a warning buzzer and light. It is different from the STI system in its packaging and control electronics. The Search-Eye system uses three sensor- detectors spaced across the width of the vehicle. An IR beam is spread horizon- tally, and vertically (though to a lesser capacity), by two long, narrow, plastic lenses. This results in a wide field of coverage for each sensor [about 1 m (39 in)], but severely shortens the range of detection to 1.1 m (3.7 ft). This system was checked in the laboratory and briefly in the field on a FEL, but its detection range was too short for use in large mine vehicles. ULTRASONIC SENSING SYSTEM Global Fabrications Co. also manufac- tures an ultrasonic-based object sens- ing system, under the trade name Sonic Radar. The principle on which the system operates involves the emission of an ul- trasonic sound burst, followed by detec- tion of a reflected energy wave returned to the source by contact with an object. The system is composed of four major parts: two front sensors, two rear sen- sors, control box, and alarm; however, the front sensors were not used in field testing because they do not affect the rear blind area for mining applications (fig. 9). 11 Electronic control box + 24 V Rear-mount active sonic detectors (4) FIGURE 9. - Schematic diagram of an ultrasonic detection system. The Sonic Radar system is installed in much the same way as the IR systems, although the sensors require more room. Also, the sensors must not be mount- ed where the engine radiator cooling airstream can flow around them, as this airflow can cause the system to false-alarm. Figures 10 and 11 show one typical location for sensors on a FEL. The system operates from a 12- or 24-V dc electrical system. The sensors are in a 14-gauge steel housing, with end brackets provided for swiveling the housing to aim the beam. Two systems were field tested, on a Caterpillar 988B loader and a Clark 275C loader. (See table 1.) 12 FIGURE 10. - Ultrasonic detection system installed on a Clark 275C. FIGURE 1 1. - Close-up of ultrasonic sensors. 13 The system was connected to energize the rear sensors when the vehicle's en- gine was switched on. The sensors will detect any object up to 4.4 m (16 ft) away from the sensor. When an object is detected, a pulsating alarm sounds in the cab. When the object is within an ad- justable range of 0.9 to 2.7 m (3 to 9 ft), the alarm will change to a constant tone. Though not specified as an option, an external solid-state alarm was con- nected to the control box, so that it would sound upon detection of an object. Both the external alarm and the in-cab alarm will sound when an object is sensed within the system's detection zone. Figure 12 depicts how the zone was mea- sured, with the black cord on the ground outlining the zone. The sonic wave coverage is actually somewhat conical near the FEL because of the separation between the right and left sensors. The coverage overlaps at 9 ft from the FEL when the transducer pairs are placed 4- 1/2 ft apart; this produces a zone of nondetection near the FEL in the area that the spreading cones do not reach. The system detects objects about 1 ft above ground level when the full cone is developed 9 ft from the FEL. Beyond 9 ft, the outline of coverage was well de- fined and rectangular in shape; overall the coverage area measured approximately 11 by 17 ft. One of the two Sonic Radar systems cur- rently being tested is on a Clark 275, and the other is on a Caterpillar 988B. At last report, both systems were working well and physically withstanding the dusty and harsh mine environment. The mine managers like the systems because of the reduction of noise from the back-up alarm, as well as the positive safety warning given upon detection of an object in the rear blind area. DOPPLER RADAR SYSTEM A short-range Doppler radar system, manufactured by Con-Serv, Inc., Omaha, NE, was recently made available for test- ing. Called an "Electronic Mirror," it is a radar device that uses the Doppler shift principle to detect the presence of a moving target within its range. The FIGURE 12. - Measuring the detection zone. 14 system is made up of a transceiver, an antenna, an intermediate frequency ampli- fier, and an audiovisual alarm. 6 The transceiver consists of a Gunn diode "Con-Serv, Inc., has also manufactured similar Doppler radar devices for use on school buses. Such a device is credited with saving the life of a child in Val- dez, AK, in February 1986, when the buzz- er alerted the bus driver to the presence of the child "retrieving a football that had rolled under the bus." The school district had contacted Russell Griffin of the Bureau regarding acquisition of this safety equipment. This lifesaving inci- dent occurred just 1 day after the sensor was installed on the bus. mounted in a waveguide cavity, providing a transmitter, local oscillator, and a barrier mixer for the receiver. Output frequency is factory preset at 10.525 GHz and the power output is 5 mW. The rest of the circuitry is card-mounted and treated with conformal coating to prevent moisture and salt corrosion. These units are mounted in an environ- mentally sealed, high-impact, plastic housing (fig. 13). A splashproof , four- conductor connecter is mounted on the rear of the unit to provide for power and connections to the audiovisual alarm and the vehicle's backup alarm. Two differ- ent types of antennas are available for use on various sized vehicles. They are constructed of diecast and machined alu- FIGURE 13. - Close-up of a Doppler radar detection system. 15 Doppler radar unit recessed in grill Extended vehicle detection zone 12-0" ■ 9'-0" 1 ll'-O" 2l'-0" FIGURE 14. - Detection zone of a Doppler radar system. aluminum and flange mounted directly to the transceiver waveguide cavity, elimi- nating the effects of noise and false re- sponses. Antenna range is adjustable to accommodate the needs of different-sized vehicles for blind area coverage. An in- tegrally molded projection on the hous- ing provides a means of mounting the unit on a universal bracket, which is then mounted on the rear of the vehicle. A Clark 275C FEL, a Caterpillar 992 FEL, and a Caterpillar 988 FEL were used during several months of testing the unit for durability and functionality. Figure 14 diagrams the area coverage in which an object can be sensed. This area will vary depending on the height , angle of declination, and sensitivity adjust- ment of the radar unit. A similar unit was also tested at the Bureau's Twin Cities facility by a con- tractor using a Caterpillar 910 FEL. Af- ter installation, the operator backed up towards an anthropomorphic dummy posed in a seated position, placed on the ground (fig. 15). The operator received a warn- ing at the 12- to 18-ft range (depending on the angle of approach) , and was able to stop in time on each trial. The range of the unit is adjustable and will easi- ly accommodate any FEL. One specially designed system was capable of detecting a person at 28 ft but was not field tested. The units have stood up well in the mine environment and operated satisfac- torily. In one case, the mine ordered a unit on its own to equip a second FEL. The feeling seems to be universal that such devices will not only protect the vehicle from rear collisions, but also, when connected to the backup alarm re- quired on the vehicle, enhance safety aspects by reducing the amount of un- necessary noise. POSSIBLE HARDWARE IMPROVEMENTS Each of the tested systems has limita- tions of width or distance coverage. In the case of the IR systems, there is a tradeoff between projection distance and width of beam with a given intensity. This is also true with the ultrasonic 16 FIGURE 15. - Testing a Doppler radar system with an anthropomorphic dummy. FIGURE 16. - Prototype synthesized voice warning syster 17 system, but sonic waves naturally dis- perse more and are difficult to project. As efficiencies in light-emitting diodes and sonic transducers are improved, along with the development of different driv- ing and receiving circuit techniques, both range and width of coverage should increase. One easily implemented improvement in any alarm system would be the substitu- tion of voice warning output for audio alarm output. Voice synthesis and dig- itizing has made great strides and continues to come down in cost. There is also evidence that voice warning is more effective in attracting attention. Such a prototype system was fabricated in the laboratory to demonstrate the applicability of the improvement. (See figure 16.) Another improvement would be to include a test switch that could power systems for preshift vehicle inspection and daily testing of the alarm. This would allow checking the system without starting the FEL and placing it in reverse gear. INSTALLATION TIPS During the course of the installation and field trials, several techniques and procedures were found to be of help in maintaining the operating systems. The following tips are listed to aid in the field installation of these systems. 1. Protect long runs of wire and cable from abrasion by installing them in conduit. If possible, it is best to fol- low existing runs of hoses, wires, or ca- bles. The liquidtite type of conduit is easy to use. Once the conduit is in place, it can be fastened with wire ties and the necessary wires and cables pulled through it with the aid of a "snake." Each end of the conduit should be finished off by using a bushing. Al- so, run wire, cable, or conduit, to allow for maintenance and removal of engine components with as little disturbance as possible. 2. Use some type of environmentally protected housing for the electronics. A NEMA 4- or 12-type box is probably the easiest to obtain. If possible, locate the control box inside the cab. 3. Physically isolate components from vibration effects if the manufacturer has not done so. 4. Locate and mount the sensors in such a way as to allow for easy re- moval for engine maintenance access. Also, make sure they cannot easily be knocked off or broken during machine operation. 5. Keep ultrasonic type sensors away from the engine cooling airstream. Mount the IR types where the least amount of dust can accumulate on them. In both cases, allow the sensor a clear view to the rear area. 6. Use an appropriately sized fuse in the power connection to the system. 7. Consult an individual familiar with the electrical wiring layout of the ve- hicle to help determine the best place to connect for primary power. DISCUSSION INFRARED SYSTEMS IR object detection systems do not per- form well in mines. False alarms can be triggered by bright sunlight and reflec- tions of the mine ground (such as white sand) . Their range depends upon the reflectivity of the detected object, which is variable due to factors like soft clothing, hard hats, reflective tape, steel machinery, and reflectors on machinery. Because of the narrow beam pattern, arrays of sensors must be employed and reflectivity within the mine must be made more uniform. Increasing the number of sensors increases the cost and system complexity yet decreases the system's reliability. The narrow beam from a sensor, mounted above the rear bumper for protection, does not have the vertical coverage to detect a per- son sitting or kneeling on the ground. Detection of this type of object would require more sensors angled downwards 18 toward the ground. The addition of re- flectors to clothing and equipment is necessary to insure uniform reflectivity in a mine. This would promote a uniform detection range, but it would be expen- sive to initiate and maintain. For these reasons, 1R technology is the least at- tractive method of detecting objects in the rear blind area. ULTRASONIC SENSING SYSTEMS Ultrasonic object detection systems utilize transverse mechanical waves at the low end of the ultrasonic spectrum. The Polaroid transducer operates with 50, 53, 57, and 60 kHz pulses and has a beam angle of 15° from the electrostatic com- bination transmitter-receiver. The Sonic Radar unit uses a separate piezoelectric transmitter and receiver operating at 32.8 kHz, with a beam angle measuring ap- proximately 34°. This system provided a nearly rectangular detection zone, but its range of only 17 ft is too short for large machines. Another problem with ultrasonic systems is that the wave velocity is considerably slower than IR light or microwaves (used with Doppler radar systems). Several waves must be sent, received, and com- pared in order to insure that the sample of wave travel time is accurate. At a velocity of 1,090 ft/s in air, a wave re- quires 1/1090 s times 2 ft, or 1.83 ms , for each foot of detection range (travel distance to and from target) . If many waves are sampled (Polaroid uses fifty- six 1-ras pulses) , the time between ini- tial detection and operator warning or system reaction time may become great enough to preclude warning the operator in time to stop the vehicle. In field tests, the Sonic Radar's sys- tem reaction time was approximately 0.5 s for an object at the edge of the detec- tion zone. This long system reaction time, coupled with a short detection range, limits the use of this type of system to smaller FEL's operating at low- er speeds. DOPPLER RADAR SYSTEMS Doppler radar systems use the Doppler frequency shift principle to detect ob- jects. This requires relative motion be- tween the system and the object being de- tected. The beam pattern is controlled by the design of the antenna. The detec- tion range is controlled by the power output, sensitivity, and shape of the an- tenna, as well as by the "radar profile," or ability to reflect microwaves, of the object to be detected. Virtually any de- tection zone range and shape can be pro- duced; however, the radar profile of ob- jects in mines is variable. In general, larger, more reflective objects can be detected at longer ranges. In-mine tests demonstrated that a system capable of de- tecting a person at a distance of 20 ft would detect a small car at 40 ft and a large metal building at several hundred feet. In most operations, detection of large objects is not a problem, as FEL's usually back away from stockpiles and mine structures such as hoppers. The Doppler radar systems are not affected by lightning, rain, fog, snow, or wind, as were the other types of systems . CONCLUSION The Bureau of Mines experimented with IR light, ultrasonic wave, and Doppler radar technology to develop a system capable of detecting an object in the rear blind area of mobile mining equip- ment. This system sounds an alarm only when a collision hazard exists, reduces the semicontinuous noise of backup alarms, and eliminates the false alarm aspect of current backup alarms. Of the three technologies, Doppler radar proved to be the best compromise because of its immunity to various weather conditions. APPENDIX.— SOURCES OF EQUIPMENT 19 Scientific Technology, Inc. 1201 San Antonio Road Mountain View, California 94043 Con-Serv, Inc. 3801 Dahlman Avenue Omaha, Nebraska 68107 Global Fabrications, Ltd 4 Twyford Road Toronto Ontario, Canada M9A1V7 Polaroid Corporation Ultrasonic Components Group 119 Windsor Street Cambridge, Massachusetts 02139 U.S. GOVERNMENT PRINTING OFFICE: 1986-605 017/40,029 INT.-BU.O F MINES,PGH.,P A. 28 27 1 U.S. Department of the Interior Bureau of Mines-Prod, and Distr. Cochrans Mill Road P.O. Box 18070 Pittsburgh. Pa. 15236 OFFICIAL BUSINESS PENALTY FOR PRIVATE USE, $300 ] Do not wish to receive this material, please remove from your mailing list, ] Address change. Please correct as indicated. AN EQUAL OPPORTUNITY EMPLOYER *>* A •- "^A A* % : %/ -+m*- t.zmx* <*> f ?£m&s\ ''++#' .v'^siw- ^.* A* V 7 ^' A ' o v-o* * t^v*'. o 4> ° - . A° , t . . , ^ A V . o » o • • • ' • ^ A* ^ ^ °°^d V ..»^Lv * -V . » • ^♦* ••!»' W #» ^ ***** -*Ste' ^^ •*$ V). ' . » * A < J ^ *' ? _?++,. \ m ■4 c> ,^ ^ ■ ^p ,<^ * ► AV y>. > ^ -. v ( 4 aV-^ *+ « V ^^S^" ^ A^° << /% 4°* 'bK ^°^ .7* A ** *: * ^ °^. >» a ^^^V r * A^ l~ .r vP - t • • r r> " A V ^ .cy . < * ■ ■ % '