Humanitarian De-mining and TRIZ (A Response to the March TRIZ Challenge)

Dr. David Levy

Editor: Dr. Michael Slocum

The TRIZ Challenge from March was on Landmines. Here is the background information provided by Ian Care:

Landmines cost less than $3 each, the new ones are predominately plastic. The automatic removal machines are only suitable for certain terrain and are expensive.

Even using relatively cheap labour (e.g. in Africa) the cost of clearing one mine is around $300 (i.e. around 100 times it's cost). The real problem is not finding mines. We can already do that quite well with metal detectors, ground radar, EM (Electro Magnetic) probes and manual prodders. Despite this the vast bulk of the £600 million or so that has been spent on "Humanitarian Demining Research" has been spent on this already solved problem, and so far has failed to produce any new tools for local deminers in the poor countries (italics added for emphasis).

The real problems are in such areas as: how to clear vegetation that may contain trip-wire mines; how to reliably survey the extent of mined areas which may be contaminated by UneXploded Ordinance (UXO) or caches of arms and not mines as such; and then how to mark them permanently in countries so poor that even painted rocks are likely to be taken because people have another use for them.

Unlike military demining, humanitarian demining is not essentially about finding mines, it is the process of being able to declare land entirely free from all explosive debris; i.e. it is about finding land where there are no mines. Frequently this does not involve any mines, it is simply making sure that a possibly mined area is not actually mined, a process sometimes called Area Reduction. Sniffing technologies are one of the most promising ways of doing this though as yet the only sniffers used in the field are dogs. There are a number of companies pursuing different routes to get the sensitivity high enough to be able to confidently declare an absence of mines.

So the challenges are: How do we mark an area that is known to be mined, or is munitions free in poor countries? How do we give poor people in poor countries the means to clear land to make it habitable and cultivatable?


Background information can be found in the book by Rae McGrath (co-founder of Mines Action Group:
"Landmines and Unexploded Ordnance -- A resource book" by Rae McGrath, Pluto Press. 2000.
ISBN 0 7453 1264 0 - It is quite technical in places.
£17.95 from <>
$79.95 from <>
The following websites of two of the leading demining NGOs (Non Governmental Organisations) have useful information and more if you follow the links they list: <> and <>
There was an article on the topic in New Scientist - Vol 64 number 2212, 13 Nov 1999, pages 52-53.


The following solutions were generated as projects of
2.997 Humanitarian De-mining
Spring 1999

Micro Digging Flail


Manual probing is a slow and dangerous process. The object of this project was to build a simple, inexpensive device that allows an operator to perform this task, at least as quickly, and hopefully more so than with the existing approach.

The machine being investigated to replace manual probing is a man-portable digging flail. The second generation of the device has a 20 cm wide flail, with three chains, powered by a small gasoline engine, in this case a modified chainsaw. See Figure 1. Not shown is the frame that catches earth, rocks, roots etc as they are thrown backwards out of the hole. The flail is attached to a long, relatively small diameter extension stick, shown at the top of Figure 1. The extension stick holds the flail over the "hot spot" (a site detected by a metal detector as a potential land mine) and keeps the device itself out of direct contact with the bulk of the blast. Air pots between the base and flail arm allow the device to lower by gravity at a controlled rate once released by the RC unit. The device weighs about 13 Kg, and is foldable for easy transportation. Washers (4 cm diameter) attached to the ends of each chain allow the little device to remove soil from the ground quite effectively.

Figure 1 - Micro Digging Flail, Prototype #2



Work to date includes two prototypes. The first prototype was electrically powered, with a wooden frame. It was used to investigate the concept of digging with a flail-type device, and to develop the digging heads located at the ends of the chains. This device proved that digging was a viable concept. The first prototype dug the hole shown in Figure 3 in 30 seconds using a 1 HP motor.

Figure 2 - Prototype #1, shown with the class (l to r) Felipe Varela, David Levy, Jeremy Walleck, Andrew Heafitz

Figure 3 - Hole Dug by Prototype #1

The second prototype, which is unfinished at the time of this writing, is built from basic materials such as steel pipe, that would be easy to find and repair in third world countries, and powered by a small gasoline engine to aid portability. It weighs 13 Kgs, which is still a little bit heavy, but has opportunities for weight reductions, such as replacing the portion of the base that contacts the ground with wood rather than steel.

The current idea is to use this micro-flail as follows: once a hot spot has been found with a metal detector, the deminer positions the flail over the spot and turns it on. (It currently works by radio control, allowing the operator to be anywhere in a 100 meter radius.) Once actuated, the device rotates and lowers itself into the ground, automatically excavating whatever was beneath it. The dirt is thrown into a contained catch area (a box, without a bottom, on the ground). If the site contained just a random bit of metal, the offending piece of metal will be easier to find in the catch area, potentially by a magnet in the loose pile of dirt that the device creates. In the event the device sets off a mine, it is designed to withstand AP blasts. (There is an assumption that if the area is known to have anti-tank mines with metal detector signatures that are indestinguishable from an AP mine, this device should not be used.) If a UXO or unexploded mine is somehow thrown from the hole, again, it will be easy to see, and would then be blown in place. Set up time is minimal. The flail that can dig out a 20 cm wide x 30 cm long x 15 deep volume in medium ground in about 30 seconds using a 1 HP motor.

Our goal is to have a cycle time that is significantly faster than conventional probing, and would also be safer because the "prober" is not working over the mine. The hope is that a machine like this could bridge the gap between mechanical mine clearing, and hand probing. Another advantage is that the device is capable of digging through soils filled with small roots and rocks, an environment where a probe is almost useless.

We perceive this project to fall into the category of “high risk/ high reward.” It is well-understood that the incorporation of this device into mine clearance operations would entail a major alteration of the SOPs, and even the basic approach to the process as a whole. In this one project, we believed that the potential for large gains in speed and safety may justify large deviations from the existing approach.

The first prototype proved the viability that a man-portable device can determine that a site is free of a land mine in half a minute using a small motor. The next task is to determine how this device could used in an effective manner in a large scale demining operation. There are significant logistical issues to be addressed. This is work that must be done with personnel familiar with the real-world limitations imposed by day-to-day field operations.

The prototype itself needs to be tested both by digging up false positives, and seeing how long it takes to isolate the piece of metal, and by exploding AP mines underneath to determine its robustness. We expect that the coupling between the explosion and the device is sufficiently small that the device will not be damaged, and that the extension stick will survive multiple blasts.

Extensive work remain, but the rewards of a semi-automated digging flail such as this could be high.


Magic Grapple

Trip wires present a dangerous nuisance in mine fields. Theoretically, a grapple may be tossed across a suspected area, dragged back and thereby pulling on trip wires, thereby detonating these booby-trapped mines in a controlled manner. In practice, grapples are almost never used for two reasons.

  1. Grapples are designed to grab things. As a result, they successfully grab trees and bushes, and as a result are nearly impossible to use in locations where vegetation is present. Often times the deminer needs to be quite close to the grapple to perform the gyrations necessary to free it.
  2. Trip wires, by their very nature, detonate land mines located in uncertain location. It is therefore critical that the deminer be located quite some distance before pulling a grapple with the intent of detonating hidden land mines.

The combination of these two problems makes grapples all but useless.



Our goal was to create a grapple that would catch trip wires, but not catch vegetation. The intent was that by eliminating the first problem, several methods become feasible to approach the second problem. Furthermore, we wanted the design to be simple, and inexpensive to manufacture.


The design could not be simpler. We decided that a sphere attached to a rope had a strong potential to meet the design objectives. We tried two different diameters, 65mm and 80mm, resulting in weights of 1Kg and 1.8 Kg respectively. In order to eliminate the risk of energy being stored in the rope while it was held in tension, a Kevlar rope was used. Because the Kevlar is essentially inextensible, it eliminates the risk of the grapple flying towards the deminer as it releases from tree limbs and the like.

Figure 4 - Two Grapple Prototypes

Both spheres are made of a mild steel. The smaller was made by drilling a ready-made sphere. The larger was machined from a solid block, and then drilled. The kevlar rope was then threaded through the hole and knotted.

Both devices were thrown numerous times into three types of vegetation to determine how effectively they avoided entrapment. These types included: low and dense shrubbery; a dense growth of trees, each approximately 3-5 meters tall, and some larger many-limbed trees. In many cases these types were combined in one area.

Trip wires were placed at different heights in the forest, from two inches off the ground to four feet high. The grapple was thrown approximately 15 to 20 meters into the forest with each throw. At no time while the grapples were being drawn back was either snagged by vegetation such that it could not be easily removed by pulling on the rope. In approximately 20 percent of the throws, the grapple was lofted by a branch or shrub so that it never contacted the trip wire. Such would be the case with any grapple. Numerous throws are required at different heights to ensure effective clearance. However, in the cases where the grapple contacted the trip wire, the trip wire was pulled with sufficient force to actuate a land mine. It required at least 20 pounds force to release the grapple from contact with a trip wire, and in many cases the grapple would not release from it until the trip wire broke.

It was determined that the smaller of the two grapples was preferred. While the heavier grapple was more effective at remaining at ground level as it was pulled back through the forest, it could not be thrown as deeply or as flat a parabola. The smaller prototype could be comfortably thrown deeper into the forest, required less force to remove

Additional work should be done to confirm develop a better statistical understanding of the effectiveness of the device. However, before this work is undertaken, there should be an exploration of whether or not a device that solves the first problem of the two noted above is sufficient to make grapples a useful device. Our observation (as researchers) is that once a grapple device may be easily pulled through vegetation it becomes much more feasible for the deminer to hide in a hole, behind a shield (such as a car) or simply at a large distance, thereby thereby creating the option for grapples to be a useful tool. This believe should be corroborated with persons in the field, to determine the validity of this belief, before additional work is done on this project.


Smart Glove and Probe

One of the early class observations was that the manual probing process is only tangentially a search for land mines. More accurately, the manual probing process is the search for bits of metal in a field that also contains a few land mines. Based on this observation we decided to spend some time ignoring the land mine problem, but instead focusing on improving the rate at which bits of metal may be found within the context of the process currently used to find land mines.


Our design objective led to a few innovative applications of traditional metal detection technologies. Two ideas seemed more interesting that the rest: 1) a metal detector built into a glove, providing the prodder with the ability to constantly reassess conditions at the hot spot throughout the process, and in some ways provide him with a “superhuman” ability in the sense that his hand itself has the ability to find metal; 2) a metal detector built into the tip of a probe, thereby revealing a completely new set of information as the manual process is performed in the traditional way.

An important consideration regarding the utility of these devices is the issue of sensitivity. Metal detectors as they are currently used for mine clearance require constant adjustment. However, this application does not have the same requirements, because it is already known that a signal exists within a few inches. The issue is only one of pin-pointing it. Therefore, it is acceptable to set the sensitivity relatively low so that only when the deminer is within a few centimeters will a signal be heard.

1) Smart Glove

Figure 1shows a glove with a metal detector built into the palm. Traditional PEMI technology may be used to provide the worker near-constant feedback with regard to the existence of a metal fragment throughout the probing/excavation process, simply by moving his open hand over the surface.


Figure 5 - Palm detector in a glove

Figure 2 shows the device itself. The device is powered by a cable that runs down the prodder’s arm to a metal detection unit mounted to his back or waist. Because the device does NOT need to be very sensitive or accurately ground balanced, the unit can be small and relatively inexpensive, such as the cigarette pack-sized coin detectors commonly sold for under $50.

Figure 6 - Palm-sized Metal Detector Prototype

2) Smart Probe

Figure 2 shows a PEMI metal-detection coil built onto the tip of a fiberglass rod, representing a probe stick. This design searches passively for the bit of metal as an active search for the mine is being conducted. Often times the deminer passes the probe past the sourceof the false positive signal in the search for a potential mine. With the smart probe, he will be alerted that he has just passed it. This knowledge dramatically shortens the overall search process.

The smart probe prototype (shown below) is obviously lacking in several respects. Obviously it is too short. More importantly, the second generation prototype will have a co-axial coil orientation, rather than co-radial, so that it may be made within the material of the probe stick, so that the electronics will be isolated from abrasion and flexure. A trough will be routed along the stick length to contain the leads, and then potted with an epoxy fill. (Note that all Smart Probe devices would be based on fiberglass sticks.)

Figure 7 - Smart Probe Prototype

The palm detector and the probe detector work well at finding buried bits of metal on the order of one gram with the sensitivity adjustmet such that constant ground balance is not necessary. However, the issue is not whether or not we can make such devices operable. The issue is whether or not prodders would benefit from having a metal detector built into their palm, or probe.


Both devices need to be developed to the stage they may be field tested.


Safe Practice Mine


It is highly desirable to have practice land mines for deminer and mine awareness training. There are several versions of such devices, but they tend to be expensive (i.e cost money). Another issue is that some practive mines may be easily modified to become real mines. Therefore, our goal was to create an inexpensive and safe practice mine that can be readily made by anyone.

The design consists of simple wooden pieces actuated by a few rubber bands. When the mine is stepped on, a flag bursts out of the ground (see Figure 1) to announce that it has been activated.

Figure 8 - Activated practice mine

Figure 9 - Practice mine above ground

Figure 10 - Before activation

Figure 11 - Interior view after activation

Figure 4 shows the rubber bands attached to the case and the flag post. The firing pin is shown to the left of the flag post. The firing pin prevents the flag post from standing up because it rests between the lower leg of the flag post (near to where the rubber band attaches) and the body of the mine. When the firing pin is forced downward the flag post is free to rotate. About 1 Kgf is required for actuation.

Figure 5 shows an exploded view. There are many variations on the theme. Our version is detailed in Figure 6.

Figure 12 - Isometric view of practice mine

Figure 13 - Parts for making a practice mine


The device works very well and can even be buried up to a few centimeters below the surface.



Go ahead and make your own.


Rotary Probe


Many mine fields have difficult probing conditions. Hard soils, roots, and rocks are common features of mine fields that make probing difficult. Reducing the force required to perform manual probing operations in these conditions should increase the speed and safety of the process.

The idea was to use an off-the-shelf device to rotate the probe stick at low speeds, such as 10-20 RPM. The objective was not to “drill” into the soil but to use an asymmetric probe (such as an oval) so that the soil was broken up to ease entry of the probe.

Figure 14 - Rotary Probe

A Black and Decker cordless screwdriver was modified to be ON constantly and mounted to a spring scale to provide force measurement during probing. These little battery-powered tools are
cylindrical in shape and cost about US $20. See Figure 1.


The device was mounted to a customized spring scale to measure insertion forces under a variety of conditions.

Figure 15 - Rotary Probe Mounted in Scale

Five probe insertion measurements were taken, each of a variety of soil types.

The rough data is as follows:

Soft Ground:

Round - 2.5 Kgf

Oval - 1.4 Kgf

Varied Oval - 1.1 Kgf

Medium Ground:

Round - 3.1 Kgf

Oval - 2.0 Kgf

Varied Oval - 1.7 Kgf

Hard Ground:

Round - 10.8 Kgf

Oval - 4.3 Kgf

Varied Oval -3.2 Kgf

Field tests are required to determine if the idea is well-received. If the concept is sound, then it would be appropriate to develop a sealed version capable of extended use in real-world conditions.

Detachable Probe Shield

To protect the hand of the prodder, some probe sticks are made with shields built into the handle. If a mine explodes during the probing process, the shield provides an extra level of protection to the hand grasping the tool. However shield-equipped probes are difficult to transport because they stack poorly. This is a significant problem because of the hard-to-access locations in which demining typically occurs. Equipment must be carried into the field each day, and portability is an important concern. To address this issue we focused on the design, optimization, and manufacture of mine probes with detachable hand shields.

The objective was to develop a probe that provides detachable hand protection that would be readily accepted by the demining community. We recognized that a simple design, one which could be inexpensively manufactured and easily used, would be of significant benefit. We imposed the requirement that our probe must not interfere with the deminer's technique. It was felt that altering the existing ergonomics might lead to rejection by the demining community and that would negate any effort to improve safety.

Two designs, one stressing simplicity and the other durability (perhaps to the point of excess) were chosen for development and field testing. The design parameters were safety, simplicity of manufacture, and ease of assembly. Two designs were selected:

Design 1) Taper Lock
A taper (determined to be one degree) is machined onto the end of the probe handle and into the mating a hole on the shield. A wood mock-up quickly confirmed that the taper lock could provide a secure connection between the shield and handle. The shield was attach from the front of the handle, so the shield will remain in place in the event of a blast. The taper has been designed excessively long so that as the device wears there will still be positive locking between the shield and probe handle. In the final design, the taper lock is surprisingly secure, and

Figure 16 - Taper Lock Probe Shield

Design 2) Threaded Fastener
The threaded fastener design offered an additional function. In this design the shield is held to the handle with a thumb screw mounted on the front face. The thumb screw is permanently attached to the shield, so that the design still has a total of two pieces. The thumb screw attaches to a metal thread deeply embedded within the handle. This design takes more time to manufacture and to assemble than the taper lock system, but has the added advantage that the deminer can push forward on the shield without any concern.

Figure 17 - Screw Lock Probe Shield


Both designs conform to Andy Smith's recommendations for probe safety and ergonomics. The recommendations cover handle and shield dimensions, probe length, probe/handle attachment, and material choice. The probe shield was cold-formed using a vice and a bending bar. Lexan was used for the handle and the shield. Both designs were equipped with the MIT tapered probe tip. Also, the handles were knurled to facilitate grasping in hot, humid environments


Both designs seem to work well. The devices disassemble quickly, and significantly improve the portability of shielded probes. However, detailed evaluation must be done by field personel.

Figure 18 - Two disassembled prototypes

Feedback is necessary from the field to establish which of the two designs is preferable. If prodders do not have a tendency or desire to use the shield as an aid to force the probe into the ground, than the taper lock design would be preferable, because it is less expensive, easier to make, and faster to assemble and disassemble. If it is determined that the deminer shield have the option of using the shield as a force point, then the thumb screw version is required and additional testing should be done to confirm that is no risk of the thumb screw being blown off in a blast.

Six additional prototypes have been manufactured and are ready to fulfill a request from Zimbabwe.

Visor Restoration Project

It makes sense to wear face protection when looking for landmines. Unfortunately visors tend to become scratched and hazy during normal field use. In order to see their work clearly, prodders often lift their visors, negating the purpose of wearing them. An effective low-cost method to restore the clarity of these visors, without degrading their mechanical properties would be useful.

Clear polycarbonate was scratched extensively with sandpaper to simulate extended use. Then various approaches were tried to restore clarity to the material. The methods fall into three broad categories.

  1. Cleaning with a solvent
  2. Solvent bonding a thin layer of clear material over the surface.
  3. Heating

1) Cleaning with Solvent

Dipping was the preferred method to apply solvent. (Wiping and spraying were also tried with poor success.) In each case, we used acetone. The samples in Figure 1 were scratched and the dipped in acetone. The upper piece was deeply scratched. Dipping resulted in marked, but not perfect improvement. The lower sample was less deeply scratched, but still enough to completely obscure the visor. This sample was restored to near-new quality with an acetone dip.

Figure 19 - Dipping Test Results

Several dozen material samples are currently in an accelerated UV degradation test chamber so that the combination effect of exposure to sunlight and acetone may be measured. These results will be available in a few weeks.

The concept is that if solvent dipping can effectively restore visors to new condition it would be cost advantageous to make a curved plastic bucket, shaped like a visor, capable of holding a fraction of a liter of solvent. Visors may then be dipped and restored to new at extremely low cost.


2) Solvent Bonding a New Layer

Similar results were found with solvent bonding an additional thin layer of material. While deeper scratches were removed with this method, bubbles were difficult to avoid. This method has the advantage of adding material over time and therefore eliminating concerns that the mechanical integrity may be compromised.

3) Heat

Heating with a torch seems to work fairly well, although raised some concerns about uneven tempering. Further work needs to be done.

This project is still in progress, although early results appear promising. More will be learned after the samples are removed from the UV accelerated life test and been tested for mechanical degradation.