Collision Avoidance Systems

Published on February 2017 | Categories: Documents | Downloads: 78 | Comments: 0 | Views: 739
of 15
Download PDF   Embed   Report

Comments

Content

Collision Avoidance Systems – CWS
Compiled By B.Arunachalam, Manager- Hospet
Extracts from Collision Avoidance Systems Technical Report April 2007 by Anglo American

General In many cases the mobile machine operator are not aware of light vehicles around their machine. Due to the size of haul trucks, the operator is not able to see significant areas around the machine. Figure shows blind spot areas around a small 50ton capacity haul truck.

Most collisions occur in congested areas – in the pit or park up areas • Light vehicles are the major at risk vehicles • Collisions with light vehicles primarily happens because the heavy vehicle driver did not see the light vehicle • Whilst heavy to heavy vehicle collisions causing injuries occur, the risk and consequence is far less than light/heavy vehicle collisions • Any system chosen must reduce the current risk. When a collision situation (route contention) has been detected the CWS system is expected to series of progressive alerts to the driver. These progressive alerts indicate that the distance between the vehicles is closing and the driver should take action. These progressive warnings can escalate from audio tone only to audio, visual, or other warning modes. Driver – vehicle interface requirements These requirements define specific ways in which the CWS systems interface with the driver (i.e., Driver-Vehicle Interface), and include indicators, displays, and warning methods. CWS should utilize different audible tones (e.g., different pitches, patterns, lengths, etc.) or tactile warnings to provide multiple progressive warnings as an object crosses the warning thresholds. The CWS system should include a visual indicator when no vehicles or objects are posing a hazard. The indication may be provided by an instrument panel warning light or an indicator that is integral to each system. The CWS system should use a visual indicator to provide system operational status. This status may be indicated by an instrument panel warning light or an indicator that is integral to each system. The CWS system should use a visual or audible indicator to indicate a system failure or malfunction. This status may be indicated by an instrument panel warning light or an indicator that is integral to the system.

The CWS system indicators should be clearly discernable in direct sunlight and at night. Consideration should be given to adjust the brightness of the indicators during night operation. Excessively bright indicators have been found to be a distraction to operators.

1.0 TYPES OF COLLISION AVOIDANCE The first and most prevalent, is close proximity (<20m) type collision avoidance systems which are suitable for slow speed applications. The second class of systems detects longer ranges (50m – 150m) and is meant for higher speed (10km/h – 50km/h) incidents. Below is a description of both classes of systems as well as typical scenarios in which they would be applicable. When deciding on an appropriate collision avoidance system, operations should consider the typical risks and past incidents associated with their operation in order to guide them in deciding which class of system to use. 1.1 Close proximity (<20m) – slow speed type collisions (0km/h ≤ Speed ≤ 20km/h) The intention of these types of systems is to warn the haul truck operator of light vehicles / pedestrians that are in close proximity to the haul truck and possibly in one of the operator’s blind spots. This is because the available systems that target this segment operate with maximum detection ranges ranging between 8m (Preco Radar) and 20m (Hazard, Booyco RFID etc). Typically these systems are designed to assist the operator at start-up (before moving the machine) or whilst reversing. The speed of 20km/h was calculated as the maximum speed under which this class of CAS will provide sufficient warning to an operator to allow him/her to avoid a collision (Table 1 and Table 2). 1.2 Stopping distances Table 1 show the calculations for stopping distances and stopping times for a loaded haul truck driving downhill on a 10% gradient. This was considered the worst case scenario and hence would represent the maximum stopping distances and maximum warning time requirements that a CAS system would need to provide

Reaction of operator tim e

D tan trav is ce elled b efore

bring h aul tru to a stop ck

Total distan need to ce ed

(in clud ingreaction time)

Tim to b g truck to a e rin

W th C llis nA o a c T c n lo yb a letow rnth ill e o io v id n e e h o g e b a e o a w in to v id p te tia c llis is a h e p ra r im eto llo o im l rin io ? ll to toe toh tima oa awh ntobo ghnm c in toafu s p llo

Safe swervingspeed

S afe s werving speed

k /h m 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 5 5

m /s m sc e 1 4 0 2 .9 0 2 8 1 8 .5 1 4 2 6 7 .1 1 5 6 1 .7 4 1 6 9 2 3 1 8 3 3 3 1 9 7 4 5 1 11 1 5 .8 8 1 15 2 7 .4 4 1 19 3 9 .3 1 1 13 11 5 1 .2 1

k /h m 5 1 0 1 5 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0

m /s 1 4 2 8 4 2 5 6 5 6 5 6 5 6 5 6 5 6 5 6 5 6

m 0 2 .8 4 .2 5 .6 6 .9 8 .3 9 .7 1 .1 1 1 .5 2 1 .9 3 1 .3 5

sc e 1 .3 4 .6 6 .0 5 .3 8 .6 9 .9 1 .3 1 1 .6 2 1 .9 3 1 .1 5 1 .6 6

m 0 .9 6 .5 1 .4 2 2 .3 0 3 .9 6 4 .7 9 6 .4 4 8 1 9 .4 9 19 1 .1 11 4 .8

P c re o Ys e Ys e N o N o N o N o N o N o N o N o N o

Hz r a ad Ys e Ys e Ys e Ys e N o N o N o N o N o N o N o

3 PPD D Ys e Ys e Ys e Ys e N o N o N o N o N o N o N o

CSA AC M Ys e Ys e Ys e Ys e Ys e Ys e N o N o N o N o N o

Ac m e c u in Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e

V ra o d Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e

Table 1.Comparison of different CAS technologies to warn operator in time (and Distance) to prevent collision by bringing truck to a complete stop.
W th c llis na o a c t c n l g b a l t wr ill e o io v id ne eh oo y e b o an e th o ea rins ff e ttim toa w im so e p r to u ic n e llo h to l w d w hsmc i eto2 k /ha dtaee aiv at n o n i ahn 0 m n k v s e cio . Itisas md a e aiv at nt k na h h r su e th t v s e cio ae t ig e s e d wl r s ltin s c n ayc ll io ( r i gi t p e s il eu a eo d r o is n divn no th b r fo eg e em r . ) S ot hr Lw o Rn e ag Fe rq Ms eh H hFe i rq g Rd r aa RID F N t ok ew r RID F GS P D ta c t ae d is ne r v lle t bin mc in o r g ah e u de c nr l n r r o to ( 0m )inod r 2 k /h re t ev = inmm o r emi u C S ann A wr i g d ta c n e e ) is n e e d d 00 . 28 . 42 . 56 . 1. 52 2. 87 4 0 5. 52 7. 22 9. 06 1 18 1.

Ln og D t ne i ac s r dr aa

SED PE k /h m 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 5 5

Dc le ee rt n aio m '2 /s -. 5 10 0 -. 5 10 0 -. 5 10 0 -. 5 10 0 -. 5 10 0 -. 5 10 0 -. 5 10 0 -. 5 10 0 -. 5 10 0 -. 5 10 0 -. 5 10 0

P eo rc Ys e Ys e Ys e Ys e N o N o N o N o N o N o N o

Hz r aad Ys e Ys e Ys e Ys e Ys e N o N o N o N o N o N o

Ms eh N t ok ew r Ys e Ys e Ys e Ys e Ys e Ys e N o N o N o N o N o

RID A F CS CM A Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e N o N o N o

GS P Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e

V RD OA Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e Ys e

For the purposes of these calculations two actions that an operator could take to avoid collisions are assumed. The first avoiding action is to bring the machine to a complete stop from a predetermined speed. The second type of avoiding action was to reduce the speed of the machine from a Predetermined speed to 20km/h. At 20km/h it was determined that the operator could safely swerve to

LongDistance Rad ar

S hort R g Rad an ar

High Freq RFID .

Low Freq. R FID

Mes Network h

brak ap es plied

com plete stop

d tance is

S PEED

S PEED

G PS

avoid a collision without loosing control of the machine and incurring secondary collisions such as driving into a berm or overturning. It can be seen from Table 1 that for both scenarios the slow speed close proximity class systems are effective up to 20km/h. Beyond this speed this class of devices will not provide adequate warning time and insufficient time to stop or slow the machine to avoid a collision. Technologies that specifically target close proximity type collisions are radar based, low frequency radio identification (RFID), and high frequency RFID. Technologies using Mesh Networking and Machine Vision Cameras are currently being developed. It should be pointed out that most long range high speed type collision avoidance technologies would also be suitable for close proximity-high speed collisions albeit at a higher capital cost. 1.3 Collision scenarios –Figures show the typical type collision scenarios.

Figure 1 Collision at intersection where one machine fails to adhere to the stop signal Figure 2. Lane departure one truck drifts into the oncoming lane. Figure 3 Rear truck moving faster and front slower

Figures 4,5, 6 &11 Haul truck reversing into parked obstacles that are in the blind areas of the operator Figure 7 . Light vehicle overtaking haul truck Figure 8 Haul truck parked or moving inside the swing radius of shovels, draglines etc. Figure 9. Haul trucks simultaneously reversing at crusher / tip or inside the pit Figure 10 Haul truck reversing towards an embankment. Figure 12. Haul truck on collision course (at slow speed) with fixed structure. Figure 13. Haul truck runs into berm. Figure 14. Truck following with differential speed. Forward machine is slowing down. There is potential for duck tailing or forward machine is parked in which case a collision is possible 1.4 Collision scenarios --:- Long range (0m – 150m) High Speed (0km/h ≤ speed ≤ 55km/h) collision warning systems The second class of collision warning system is high speed-long range collision avoidance systems. These systems operate on the principle of detecting potential threats (light vehicles, other haul trucks etc.) in the direction of travel and providing sufficient warning time to the operator to take action. Action could include applying brakes to bring the machine to a stop or it may involve slowing the machine to a controllable safe speed before swerving to avoid a collision. Table 1 shows the calculations of stopping distances and times. It can be seen that this class of CAS that these systems would operate at speeds up to 55km/h where stopping distances of 150m are required on a downhill gradient High frequency RFID which typically has a range of 50m – 70m would not provide sufficient warning time above 35km/h. At 35km/h the operator needs approximately 65m to bring the machine to a stop. Hence the 50m warning is insufficient. If it were possible for the operator to slow the machine down to 20km/h and take action then high frequency RFID would be effective up to a speed to 40km/h Long distance radar and GPS have long ranges. Long distance radar can detect up to 150m while GPS based systems are effectively limited to 500m by the vehicle to vehicle wireless communications network. Both these technologies would be able to provide adequate warning to the operator at all speeds. Lane departure one truck drifts into the oncoming lane . A truck drifts onto the oncoming lane. Several factors can cause this including, driving around a curve, loss of concentration due to operator fatigue, avoiding a slow or parked vehicle ahead etc. A long range CAS would detect a potential collision and provide warning to the operator to correct course Slow moving machines or parked machines pose a potential hazard. This can be even more hazardous if the parked machine is around a curve. A long range CAS system should be able to detect slower moving or parked machines in time to allow the fast moving machine behind to slow down or stop. Haul truck runs into berm. Several factors can lead to a truck running into a berm. Taking avoiding action, operator fatigue, poor road conditions are some of them. Technologies that use tags or receivers (RFID or GPS) will not be effective in preventing collisions with embankments.

Technologies are available to detect if the operator is driving within his lane and alarm the operator if he drifts out (Accumine, Dephi systems)
2.0 COLLISION WARNING TECHNOLOGIES

These technologies include Radar, Radio frequency Identification, Mesh Networking, GPS and Machine Vision systems. The aim is to describe the operation of these technologies, discuss their advantages and limitations that are applicable to surface mining conditions. 2.1.1 Radar There are several radar based CAS systems. Fitted only to the host unit but can detect other people, objects and equipment Works on line of site although some technologies use multiple units and software to “bend” the signal slightly to try to avoid spurious alarms Antennas fitted outside the vehicle with alarm units (rows of lights or pie chart screen for visual and audible alarm) fitted inside cab. 2.1.2 Basic principle of radar (RAdio Detection And Ranging) The basic principle behind radar is a signal is transmitted, it bounces off an object and it is later received by some type of receiver. Radars use certain kinds of electromagnetic waves called radio waves and microwaves. EM waves transport energy through a vacuum – this implies that the speed of the signal is constant (c = 292,792,458 m/s). This feature is used to determine distance calculations to targets and is called ranging.. Once the radar receives the returned signal, it calculates useful information from it such as the time taken for it to be received, the strength of the returned signal, or the change in frequency of the signal. This information is then translated to reveal useful data: an image, a position or distance away and the velocity (speed) of a haul truck. 2.1.3 Radar for mining equipment There are several types of radar systems which include Pulsed, Synthetic Aperture Radio, Doppler, Continuous Wave etc. Collision avoidance systems for mining equipment that use radar as their primary technology have the advantage that they are low cost.. Another advantage of radar its ability to work in all weather conditions (rain, mist, fog etc.) – although significant mud build up onto an antennae can cause operation to deteriorate. Radar systems are relatively easy to install. The positioning of the antennas is critical. If the units are installed too low in front or behind then too many rocks etc. will be detected. Mounting the unit too high may cause the radar to miss low objects which, depending on the size of the haul truck, might be light vehicles or people. Radar systems can detect any object that is able to reflect the EM signal back. Hence other vehicles, people, rocks, buildings, trees etc will be detected. The detection of rocks and foliage may be seen as a false alarm. Too many false alarms could be seen as a nuisance to operators. Nuisance alarms are those emanating from objects of which the operator is already aware or from objects that pose no danger. Too many nuisance alarms may result in an operator not taking a system seriously and ignoring alarms even when a potential collision is imminent. Because of the potential nuisance alarms. In addition it is advisable to integrate radar systems with cameras in order to minimize the number of LCD displays inside the machine. The forward and rear cameras should be activated by a switch on the transmission and the radar alarms should be superimposed on the picture. 2.2 Radio Frequency Identification (RFID)

2.2.1 How it works Radio Frequency Identification operates on a Reader and Tag principle. In its simplest form RFID systems work on the principle of inductive coupling. The reader's antenna coil generates a strong, electro-magnetic field, which then penetrates the tag. The tag is essentially a coil – many windings of copper. The electromagnetic field induces a voltage that allows the tag to function. The tag or more accurately called the transponder sends back an electromagnetic signal to the reader. Each tag normally has a unique identity, so when the reader receives the signal from the Tag it can identify where it came from. RFID systems can be classified as active or passive, high frequency or low frequency.. These features are important when evaluating RFID systems for collision avoidance, because they determine the range (read distance) of the system as well as the effect of the metals etc on the system. 2.2.2 Nautilus Buddy Haul Truck System Nautilus International has a low frequency RFID product called the Nautilus Buddy Haul Truck System. It is 125 kHz system that uses two loop antennae’s to propagate a magnetic signal. The system uses a radio frequency communication link to transmit signal back to the haul truck. One Antenna is mounted at the front of the truck close to the front railing and a second at the rear. These antennas radiate a field which surrounds the truck completely, including underneath the truck. A tag (Belt pack) is mounted on people or on light vehicles. The range of the coverage is shown where it can be seen that it fluctuates between 15m in the front of the truck to approximately 9m on the sides and the rear of the machine. The complete Buddy system CAS equipment is shown below .

The system calculates the distance that the potential threatening light vehicle/ haul truck is, and displays this information on the haul truck display as well as on the Tag inside the light vehicle. The system uses two antennas to differentiate if the target is approaching from behind or in front. The system has an optional Cab Video Unit where the data can also be displayed This display data includes warning messages for the driver and also shows whether the "target" is a pedestrian or a small vehicle. An audible alarm built into the Cab Video Unit will increase in frequency as the Haul truck approaches the "target".

Once the Haul truck gets too close to the "target" the Haul truck's air horn will automatically sound and the headlights will flash on and off to warn personnel and small vehicles in close proximity to move away to a safe distance. Summary The Nautilus system has excellent coverage all around the machine and even underneath it making it suitable for close proximity slow speed collisions. On the other hand the system is expensive to install and every haul truck and light vehicle needs to be equipped with the system for it to be effective. The size of the Belt pack could also be a deterrent for adoption. Also against this system is the relatively high cost of the unit. 2.3 High Frequency RFID systems Advanced Mining Technologies (AMT) CAS-CAM System description Radio Frequency Identification • This technology uses radio frequency signals and detection via antennas. • Requires line of sight between transmitters and receivers. • Requires each unit or object to be fitted with the technology. • Is often combined with camera technology with some manufacturers using several units to add sophistication to the alarm logic that can be programmed, and switching of cameras looking at the side of the equipment where the potential collision has been detected. • On smaller vehicles and equipment the RF unit is combined in rotary flashing light with just an alarm box in the cab • Can be programmed to alarm at set distances, and even identify specific vehicles with the alarm • Is the same technology used in modern hospitals to keep track of patients, and in factories to keep track of tools and components. The AMT CAS-CAM system is a high frequency RFID system. It operates on the 433MHz frequency which puts it into the high frequency range.. the tags provide their own power and are not dependant on the EM field emitted from the reader in order to communicate. The CAS-CAM system consists of readers, tags and cameras. Tags are installed on each person, light vehicle, heavy vehicle and other items of value. Figure shows the RF unit as well as the camera installed on the rear of a haul truck.. Heavy Vehicles also have video cameras and an LCD video display unit. The images from the camera and the alarms from the RF units are displayed, simultaneously on the LCD unit inside the haul truck.

AMT CAS-CAM RFID reader and camera. The Radio Frequency system transmits digitally coded data such as tag identification number, tag type, vehicle status and tag status. This allows the system to discriminate between classes of objects e.g. Heavy Vehicle (HV), Light Vehicle (LV), Stationary Object (SO) Personnel Tag (PT), Test Station (TS) etc. The system comes in three options: 1. Video Only system for enhanced vision applications: CAS-CAM 2. RF Only system for automatic object detection: CAS-RF 3. Video & RF systems (vision & object detection): CAS-CAM/RF® 2.4 GPS Based systems 2.4.1 How GPS works The Global Positioning System (GPS) consists of a constellation of 27 Earth-orbiting satellites (24 in operation and three extras in case one fails). These satellites are in a fixed orbit approximately 20,000km above earth.. A GPS receiver, such as a handheld unit or an electronic circuit board type receives radio messages from these satellites. The receiver then calculates its distance from the satellite sending it a message. The receiver obtains messages from at least three satellites in order to triangulate its position on the earth’s surface. Typically a receiver would use around 6 to 12 satellites (if available) to triangulate its position. The more satellites available the more accurate the position fix. Because the earth is rotating, satellites appear and disappear over the horizon, hence the more satellites that are available not only increases the accuracy but also reduces the risk of loosing accuracy. GPS Technology • Each vehicle or machine is fitted with a GPS unit • The individual vehicle or machine positions as determined by GPS are transmitted to a central computer and are scanned for potential collisions • If a potential collision is detected warnings are sent to the vehicles or machines involved using radio transmission 2.4.2 GPS based collision avoidance GPS by itself cannot do collision avoidance – it is simply a means to establish position. By knowing the positions of vehicles in a mine and communicating these positions to machines it is possible to start to have the first part of a collision avoidance system based on GPS positioning. The principle of a GPS based collision avoidance system is as follows: - Each machine is equipped with a GPS receiver to obtain its position - In addition a communication system is required on each machine in order to broadcast its position as well as “listen” to the position of other machines in the vicinity (0 – 500m range) around. The communication network is probably where the most differentiation occurs with GPS based systems. - On each machine calculations are performed to see if any machine (light, heavy etc. is a threat and an alarm is activated if it is a threat. 2.4.3 Acumine Proximity Detection System. The Proximity Warning System is a GPS based system that operates on Haul trucks, Light vehicles as well as People. Three modules are used: Haul Truck Proximity System (HTPS), Light Vehicle Proximity System (LVPS), Personnel Proximity System (PPS) and a Base Station. The HTPS alarms the haul truck driver when another truck, a light utility vehicle or personnel is within the defined proximity of the haul truck. The haul truck forms an ad-hoc mesh network with these agents, all of which are equipped with GPS, and broadcasts its

position and velocity.. The HTPS will generate a different alarm according to the threat level, e.g. truck approaching in front, vehicle behind etc. The system uses a dedicated on-board computer for processing and alarming in the haul truck and light vehicles. A Personal Digital Assistant (PDA) is used to warn personnel such as pedestrians etc. Each agent uses a GPS sensor and an Omni directional antenna for wide area coverage. All these agents are registered in a single ad-hoc network. The area of operation is by line of sight of the agents in the proximity and the area of detection. This is possibly a limitation of the system .Systems based on line of sight will typically give ranges of 100m – 500m depending on conditions. However if line of sight is obscured such as on ramped curved roads or possibly at intersections this could become an issue and the system would not detect with required level of fidelity The operator interface is either a simple audible alarm or a full graphical interface as shown in Figure

Figure Haul truck graphical interface. The system can be classed as a long range high speed system There are two major drawbacks of this system. Both relate to the wireless network system that they have adopted. One drawback involves the issue of line of sight which cannot be guaranteed in mining environments. The other involves the use of a base station which is a weak link that compromises the system long term operation. The entire fleet would be left without a collision warning system if the base station went down. 2.6 Camera Based Technologies Collision Avoidance – Cameras • A reliable technology of video cameras that the operator can select to give better vision in the blind spots of the vehicle. • A passive system relying on the operator to choose the camera. (Can be tied into the reverse gear signal or other functions) • Multiple cameras (blind side, reversing, etc.) are sometimes fitted. Additional cameras, whilst increasing visibility, can also complicate use. • Modern LCD colour screens are much more reliable than old CRT ones. When correctly trained operators can choose night settings to reduce night time glare inside the cab. • Installation of cameras will not guarantee that collisions will be avoided. However, their installation gives operators better tools to enable them to do their jobs safely. 2.6.1 Caterpillar WAVS WAVS stands for Work Area Vision System.. The WAVS system comes in four configurations – single camera, 2 and 3 cameras and a special 793 haul truck configuration. Single camera systems are popular with smaller machines such as wheeled loaders, etc. For haul trucks multiple camera units are required. The multiple camera units are automatic

units i.e. the rear camera is activated via the transmission. Side cameras are manually activated. A useful feature is that the operator cannot defeat the cameras (switch them off) as there is no manual override. This implies that the cameras are always functional. Cameras come in two configurations 115 degrees and 78 degree options which are used depending on the application. The cameras come with special mounting brackets and have internal heaters to prevent misting in cold weather. The cameras can take vibration and is sealed (High pressure washer compatible). This makes the WAVS an extremely rugged unit. In comparison the CAS-CAM cameras (discussed above) the CAT cameras can go down to 0.5 Lux (CAS CAM goes down to 1 LUX).

. It is recommended that multiple cameras be installed on haul trucks a. Four cameras (front, rear and two side cameras) are recommended. Depending on the size of the truck the side camera on the operator cab side of the truck may be omitted. Installing an additional camera under the truck could help the operator see large rocks which could damage tyres. b. The rear and forward cameras of the system should be linked to the transmission of the machine and activated automatically when the reverse or forward gears are selected. 2.6 Machine Vision Systems Machine Vision commonly known as Vision Systems is a software technique that analyses images and outputs characteristics found such as objects, characters, edges, distances, colours etc. The software does this by analyzing pixels and calculating features inherent in the pixel configuration. For example objects can be recognized using a technique called pattern recognition. In pattern recognition the system is trained using sample photographs or images. Similarly character recognition is performed. Bar coding is one of the most common forms of machine vision systems. In Bar coding a laser line is shone on the bar code to illuminate the bar. The machine recognition software detects the edges of the bar code and calculates the width of each bar. A combination of different widths constitutes a bar code. Machine vision systems are very dependant on lighting conditions. Bright light can blur edges and this confuses machine vision software. Therefore infra red light is often used to illuminate objects. Nevertheless ambient light still poses a significant problem and needs to be dealt with in an application. Camera vibration is another threat to machine vision applications and can blur images. Perhaps the most significant problem with pure machine vision systems from a collision avoidance perspective is that it is a 2 dimensional technology. This implies that standard machine vision system will not be able to decipher how far away a threat is from the host vehicle. Solution to this problem – Stereo Vision. This involves two cameras which use disparity calculations to calculate distance.

2.7 Pros and cons of various systems GPS & RF Pros – Provide an active warning – Can be integrated with camera systems – Can display multiple items – Easily configurable for many different scenarios – Reliable components – Doesn’t differentiate between day/night – Short and long range configurations – Can provide speed, distance and azimuth information – No licensing requirements – Can identify individual unit numbers – Can define zones in mine eg high risk zones, speed limits, etc – Can integrate with machine
Cameras

Cons – Not global coverage – Susceptible to shadowing effects – Not Commercially available – Requires lots of repeaters – Every unit must be tagged – Large objects can shadow smaller objects – Integrity of the systems relies on all components in series – Requires additional infra-structure on site for operation

Pros Allows to view blind spots – Provides positive ID of objects – Commercially available – Easily configurable for many different scenarios (position, units, link to gears, etc) – Reliable – Can provide day/night coverage – Operator/community acceptance – Provides clear lifelike image (orientation is same as what is seen in the mirrors) – Low cost of implementation – Improve operator visibility during manoeuvring – Provide operational as well as safety benefits – Stand alone system per EME RFID

• Cons – Line of sight detection – Cameras cannot provide an alarm – Cameras may be effected by environmental conditions – Trade of between image clarity and field of view – Distance perception is difficult – Not beneficial for high speed applications – Relies on operator to look at the Screen

Pros Identifies tagged items – No false alarms – Wide azimuth at front & rear of vehicle does not generate false alarms – Not impacted by environmental conditions – Provide an active warning – Commercially available – Can link to camera systems – Can display multiple tagged items

Cons Easily configurable for many different scenarios – Reliable? – Display can be integrated with camera systems – Doesn’t differentiate between day/night •– Does not identify untagged items – Requires management discipline to ensure all items are tagged – Cannot pin point exact position of tagged item (general direction only) – Wont detect whether tag is moving or fixed – Requires licensing to meet country laws – Line of sight detection for both units – Global support not currently Available

RFID + Cameras Pros – Identifies tagged items – No false alarms – Wide azimuth at front & rear of vehicle does not generate false alarms – Not impacted by environmental conditions – Provide an active warning – Commercially available – Can link to camera systems – Can display multiple tagged items – Easily configurable for many different scenarios – Reliable? – Display can be integrated with camera systems – Doesn’t differentiate Radar • Pros Not impacted by environmental conditions

• Cons – Does not identify untagged items – Requires management discipline to ensure all items are tagged – Cannot pin point exact position of tagged item (general direction only) – Wont detect whether tag is moving or fixed – Requires licensing to meet country laws – Line of sight detection for both units – Global support not available – Cameras may be effected by environmental conditions (Note Black and White cameras much better for night use)

Cons Requires line of sight detection

– Provide an active warning – Commercially available – Can be integrated with camera systems – Can display multiple items – Easily configurable for many different scenarios – Reliable – Doesn’t differentiate between day/night – Detects all items – Stand alone – Short and long range configurations – Can provide speed, distance and azimuth information – No licensing requirements – Can focus detection area to a defined width (eg width of vehicle for forward travel)

– Large objects can shadow smaller objects – Non-essential alarms - detects items outside of vehicles travel path? – Cannot differentiate for the type of vehicle, object or person of similar sizes – No positive unit ID on object Detected

CONCLUSION The recommendations for Anglo American operations are as follows: 1. Each operation needs to undertake a risk assessment in order to determine its risk profile. This should involve studying past collisions, near misses or high potential incidents etc. The risk assessment should also identify what actions needs to be taken in order to reduce the risk of collisions between people, mobile machinery and other dangerous machinery. It should also take into account the possible speed of vehicles and determine the minimum and maximum distances from a vehicle at which the detection and warning alarm are needed to sound off; 2. It is prudent to inform operations that all collision avoidance (CAS) systems / technologies investigated by the working group were found to have certain technical and operational limitations. This report provides the necessary technical information discussing limitations and strengths of each system. Understanding the risk profile of the mine together with the technical and operational limitations of the CAS system is a key to mitigating the risk of collisions. 3. Systems that are aimed to be used for anti collision and proximity warning should be not be named: “Safety Systems”, rather they should be referred to as “Operator Enhancement Systems”. This would ensure that the responsibility remains on individuals for ensuring their own safety. It also ensures that individuals do not rely solely on the technology for protection. Surface Mining Equipment 4. It is recommended that multiple cameras be installed on haul trucks. a. Four cameras (front, rear and two side cameras) are recommended. Depending on the size of the truck the side camera on the operator cab side of the truck may be omitted. Installing an additional camera under the truck might help increasing tyre life. b. The rear and forward cameras of the system should be linked to the transmission of the machine and activated automatically when the reverse or forward gears are selected.

c. The working group did not comprehensively evaluate all vendors of camera systems. The robustness of the camera system as well as the ability to clean the lenses is important factors to consider when making a choice of camera system. 5. In addition to cameras a RFID based system and/or a radar based system should be installed. a. If the risk assessment shows that slow speed, close proximity collisions are the major threat then radar systems should be installed. b The Preview Radar* system should be installed. The system has a limitation of an 8m range and this should be carefully considered before deciding to install the system. A recommended configuration is discussed in the body of the report. c. If the risk assessment shows that high speed, long range collisions are the major threat then an RFID based system is recommended 6 The range of the system chosen should be set to a minimum of 50m. If possible the maximum range of 100m should be used. 7 The ability of the system to detect near / alongside metal structures such as conveyor belts should be tested. If loss of detection around metal structures is found then alternative safety procedures should be put in place around these areas. 8 A reverse / back up audible alarm should be used for warning while reversing. 9. It is recommended that a reverse / back up audible alarm should be used for warning while reversing.
Collision Avoidance Technical Report Author(s) H. Faul; M Ruplal (ATD) O. Munoz; E. Riffo (Anglo Base) S. Niven (Anglo Coal) A. Naidoo (Anglo Ferrous) D. Janicijevic; V. Nhlapo (Anglo Platinum)

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close