MgM (Menschen Gegen Minen) is a German NGO that has been involved in mine clearance since 1996. Its area of operation has been limited to areas north of Luanda in Angola. Despite being a rather small organisation MgM has managed to become one of the best known NGOs in the global demining theatre. This is achieved through professional marketing and inventive thinking on how to increase productivity and reduce costs of operations. MgM uses a concept involving a variety of complimentary machinery to achieve its demining goals while the number of employees has been kept to an absolute minimum.

Initially MgM developed the MaM system which consisted of a small manual mine clearance component supported by a vegetation cutter (mulcher) mounted on a mine proof Wolf(1). The aim was to increase the speed and safety of demining by cutting the vegetation (and possible tripwires) mechanically before areas were cleared manually. MgM’s main priority has been to demine and rehabilitate the secondary road network in the northern Bengo province. A mine protected grader was therefore integrated as part of the MaM system approach. The grader would grade a road while the manual mine clearance team would visually inspect the road surface and clear the berms built up by the grader. The system was effective but it could only be applied on roads with a low density of mines. Dogs were therefore introduced to work behind the armoured vegetation cutters and as a means of quality control after grading and manual clearance. According to MgM the system worked so well that they decided to name it the Voodoo system. The principle is similar to what is used by several other organisations. High clearance rates (99,6%) are achieved through a series of complimentary approaches. The main difference between MgM and other organisations is that MgM has developed and built most of its own machinery. Old machines have been bought from scrap yards in Namibia and re-built to satisfy MgMs requirements.

The MgM ROTAR was developed as a result of practical problems experienced in the field. Mines and UXO had been found in domestic garbage and many mine suspected areas had been littered with tin cans and other shrapnel. The above mentioned demining approach had not provided a satisfactory result and MgM was looking for a new way of dealing with this problem. The desired solution was to construct a relatively lightly armoured sifting system that could safely remove the topsoil for sifting and subsequent visual inspection of all larger objects outside the hazardous area. The rotating sifter had previously been considered as a possible solution to the mine problem along the electricity pylons in Namibia. The idea was rejected. MgM, however, saw a potential in the use of such a system. After some two years of investigation and preparation the MgM ROTAR was finally developed in Namibia.

MgM sees a potential for an extended use of the ROTAR beyond the clearance of domestic garbage. Previous clearance of electricity pylons in Namibia has deteriorated the mine problem since mines have been deep buried into berms of soil or sand instead of being removed or destroyed. A relatively straight forward clearance task has been turned into a difficult challenge due to the depth of the remaining mines. It seams that sifting of the soil/sand is the best approach and MgM believes that their ROTAR is the most suited machine for the task. A second potential for the machine according to MgM is the clearance of regular mine fields and rice paddies.

The ROTAR consists of an armoured front loader with a rotating sifter bucket mounted to the front end. The front loader is a refurbished and armoured Caterpillar 916. The cabin is mine proof and all vital areas have been protected with armoured plates. Like all other front loaders the Caterpillar 916 is hinged at the middle which makes the machine easy to manoeuvre in restricted areas. Anti-tank mines would probably inflict serious damage to the machine. The ROTOR is therefore only intended to be used in areas where no anti-tank mines are suspected.

The cabin is made from 6 mm ballistic steel plates with bullet/fragmentation proof windows on all fire sides. Because of the armouring and sealing, it has been necessary to install an air condition inside the cabin. The cabin is also fitted with HF, VHF and HF PGS radio communication equipment.

The rotor is a Dutch building rubble sifter (ROTAR) which has been customised according to MgMs own specifications. Extra protection is added on all susceptible parts. The strength of the retaining bars has been increased and the grid size in the sifter is 4,5 x 4,5 cm, which means that most anti-personnel mine types will be restricted from slipping through the GRIDS during rotation and sifting. The rotor is a closable bucket with 10 teeth at the front for easier penetration into hard ground. The upper part of the bucket can be locked and opened hydraulically. After the lower bucket has been pushed into the ground and filled with soil, the upper part of the bucket will close on the lower part. The lower bucket part has a 2 mm sheet plate inside while the upper part of the bucket has a cross-ruled sieve. The rotor can be lifted and lowered in the same way as a normal front loader bucket. In addition, when closed, the rotor can be repeatedly rotated which will improve the sifting ability of the machine.

The ROTAR is normally supervised during work through a mine proof Wolf equipped with radio and surveillance camera. The surveillance camera is fitted to a high telescopic mast on the top of the vehicle. The camera has limited zooming abilities, which makes it difficult to supervise and direct the work from inside the Wolf. During the tests, the ROTAR was therefore not supervised from the Wolf but rather by a Supervisor who walked freely around the whole test area. During real clearance this would not be possible due to the blast/fragmentation threat from mines/UXO. However, MgM informed the evaluation team that they intended to build a portable test cabin that could be moved easily around. The test cabin would be placed on the ground near the clearance site for better supervision and control of the clearance.

On one occasion a mine jammed while the rotor tipped its contents onto the truck body. The MgM crew did not discover this. The ROTAR backed out from the truck and manoeuvred around and finally back into the minefield with the mine loosely resting on one of the bucket teeth. Luckily the mine did not fall out during the manoeuvring of the ROTAR and was finally pushed well into the bucket on the next excavation. The mine could, however, easily have fallen out into areas considered as safe without anybody noticing it. The mine was also not discovered by the Supervisor. Upon inspection a small elevation was discovered in the joint between the inner sheet steel plate and the bucket teeth. The mine had probably jammed in this joint.

Site layout
The second clearance test was carried out on flattish dry ground with low grass and few rocks. The surface was relatively hard due to natural compaction. The test was originally meant to be undertaken on a 5 x 10 meters area with 100 mines. However, due to the length of the first test, it was anticipated that it was too late in the day for the machine to clear 50 m². 50 mines were therefore buried to different depths, from 0 to 30 cm. All the mines were laid in a 5x5 metres square, which was marked with red paint. This time the mines were not laid according to their numbers but in a random way.

Supervision
The supervision was carried out in the same way as for the first test. It was agreed in beforehand that that the Supervisor and the driver should not be informed about the number of mines found or not found at any stage during the test. Unfortunately, spectators gave information about the found number of mines on a number of occasions. Especially towards the end of the test, spectators constantly informed the Supervisor about the number of missing mines. Again this reduced the credibility of the test as it may have prompted the Supervisor to work differently if he did not know about the number of missing mines.

Test results
The test was terminated after 52 minutes because of twilight, which made it too difficult to supervise the work of the ROTOR. By that time the whole area had been cleared but the machine had pushed soil into berms outside the mined area. 48 out of 50 mines had been found which gives a clearance rate of 96%. It is, however, possible that the machine would have found one or both of the two remaining mines if it had been allowed to sift longer.

Assuming that the machine would eventually have managed to remove all the mines, this would probably have taken some time, 20 — 30 minutes is anticipated. Although 25 sqm cleared in approximately 1 hour and 15 minutes is significantly faster than the first test, it remains a fact that the ROTAR system is slow and time consuming if applied as demonstrated during the tests in Namibia.

Results
Inspection of the bucket after the detonation of the mine proved that the rotor bucket and the machine had taken minimal damage. A small bulge could be observed on the inner sheet steel plate on the rotor. Metal fragmentation had also caused several small grooves in the steel side plates inside the rotor. No other damage could be observed on the machine. The rotating and closing functions on the rotor were subsequently tested and showed no sign of being damaged.

During previous tests an R2M2 mine detonation had caused a bulge and a small hole in the sheet steel plate. Other R2M2 mine detonations inside the rotor during the same test did not cause any damage. It is therefore likely that the particular mine that caused the damage was detonated while being in contact with the steel sheet plate while the other mines detonated as a result of pressure caused by the soil or rocks. The J69 used for this test was not placed in contact with the rotor. It is likely that the bucket would have taken more damage if the mine had been placed in direct contact with the bucket. However, even if the bucket had taken more damage, it would probably have been easy and little costly to repair.

Despite some minor hydraulic problems the machine appeared to be well constructed, sustainable and fully functional. No other problems occurred with the machinery, the hydraulic system or the rotor during the tests.

• The bucket appeared too small, which resulted in a limited intake of soil and easy creation of side and front berms. It is questionable whether a rotating sifter is faster than e.g. a normal mine protected front loader with an external sifter.
• Much of the sand/soil actually fell back onto the ground before the machine could move out of the mine field. Since the only way to be sure of total clearance is to remove all the soil, soil that had already been sifted but had fallen back into the field was sifted again. The machine is therefore less effective in dry areas where the soil is loose or in sandy areas. It might be more effective in clogged soil.
• The problem of safe supervision remains unsolved. During real clearance the Supervisors would not have the freedom to move around as was done during these tests. A more restrictive use of the Supervisor may well have less favourable test results.

• The rotation of the rotator and the sifting of soil created significant visibility problems. Due to the close supervision (1 — 2 metres from the excavation site), this did not cause any problems during these tests. It is, however, envisaged that dust may cause significant problems during real clearance. Under dry conditions wind is probably a precondition for safe clearance.
• Although the rotor has proven to withstand detonations from fragmentation mines such as the Valmara 69, it is not known what damage the detonation of UXO, such as 82 mm mortars would cause. UXO are found in any theatre and must be taken into consideration. It is also likely that some UXO would detonate inside the rotor due to the rough processing during the sifting process.
• Like all wheel driven machines, the ROTAR will experience accessibility and manoeuvrability problems in areas with wet and muddy soil. The clearance e.g. rice paddies is probably not possible during the rainy season, only when the paddies are dry.
• The machine proved to be able to deal with limited vegetation without any problems. In areas with thick vegetation, however, it is likely that the bush will cause problems for successful rotation. Cutting of vegetation by using a vegetation cutter is probably required.
• Although the machine managed to remove a satisfactory number of mines when taken into consideration that it will be used as part of a system approach, the clearance efficiency must been seen in correlation with the time spent for the clearance. 4 hours and 16 minutes for the clearance of 50 sqm is far from satisfactory. In comparison two manual demining pairs (two metal detectors in work) would probably have used the same time to clear the area. However, further experience with the machine coupled with the establishment of suitable SOPs and clearance systems will probably increase the speed of future clearance.
• In areas with very rocky soil it might be difficult for the EOD Specialist to recognise mines between all the rocks. Especially if the machine is working in wet packed soil where mines can not easily be distinguished from normal rocks due to the typical soil layer covering the mines. The removal of rocks for better view beneath can be a dangerous exercise, which will jeopardise the safety for the EOD Specialist.
• The grids in the sifter are given by MgM to measure 45 x 45 mm. This will effectively prevent all known mine types from falling through the grids when sifting. There are a few mines with a height lesser than 45 mm. All of these mines, however, have a diameter larger than 45 mm which will effectively prevent them from falling through the sifter. 45 x 45 mm openings is therefore the most practical grid size.

The machine would probably also have a much greater capacity if the bucket itself was larger. The inner plate of sheet steel did not prevent soil from falling when lifting the bucket. This problem can possibly be overcome by increasing the size of the plate so that it covers a larger inner area. This will probably make the sifting process more time consuming. The increase sifting time must be balanced with the decreased excavation time.

If the machine is to achieve satisfactory clearance rates, it is imperative that the work is closely monitored and supervised. The supervisor must be able to stand fairly close to the bucket when it is excavating. Ideally the supervisor must be able to move relatively freely around the machine during operations. This can not be done without the use of a mine protected vehicle. MgM should investigate the use of a small mine protected vehicle for this purpose. A hand portable protection unit on the ground will probably not be the best solution as it limits the supervisor to walk/stand on land that are not suspected as mined. In a real theatre very few areas can be considered as safe. This problem can, however, be overcome by providing safe access using a wolf with steel wheels. A second alternative is to further develop the remote monitoring system through improved video cameras. Several cameras set up on both sides of the field in safe areas might be a possible solution.

The machine did not satisfy the UN requirement (99,6% of all mines) of clearance in any of the two tests. The clearance results where, however, more than satisfactory taken into consideration that the machine is intended to be used as part of a system approach. This achievement must, however, be seen in correlation with time used for the clearance. If the machine had not worked so slowly this might have affected the overall results and left more mines behind.

The machine was able to withstand the fragmentation and blast effect from a J69 (Valmara 69) bounding fragmentation mine containing 597 g of explosive plus approximately 100 gram of plastic explosive, which was used as a primer during the test. The detonation only caused minor damage to the inner steel plate in the rotor. The inner plate can easily be repaired or replaced if damaged although this would not be necessary after this blast test.

Safety is an important issue for MgM. The ROTAR concept provides a high level of safety for the operator. The level of protection on the ROTAR and the truck tipper is very high. It remains to be seen what will happen if the ROTOR detonates an anti-tank mine with the rotor or one of the wheels. Most likely the machine will take substantial damage while the operator will be safe if properly strapped. Since the machine is not intended for use in areas where anti-tank mines cause a treat, the conclusion is that the system provides a high level of safety for people involved. This is if MgM finds a solution to the protection of the supervisor.

The machine is able to clear mines to a satisfactory level with a high level of safety for the operators. The concept is, however, slow which will make overall demining less cost effective for typical mine clearance of mine fields on flat ground. The machine may still contribute to increased overall cost efficiency if it is used to clear areas where other demining methods have little application, such as domestic garbage, deep buried mines and berms.