Andy Smith
MIne-action specialist

Manual demining tools

Magnets held on tools

The use of powerful clip-on magnets to reduce fragmentation during metal-detector based clearance was discussed on the demining forum. The discussion also covered the use of rakes with magnets attached to rakes as a way of reducing the number of metal-detector investigations that had to be made in an area. The exchanges are summarised below.

See also: Magnet on blast resistant trowel and Magnets for fragment reduction. For safety, see also the entry on the Type 72b mine.

Subject: The attraction of Rakes

It's rare that something new arises. When it does, it is almost always an obvious and simple combination of things that are already known. Competition for the kudos associated with any development is such that the only way of preventing someone claiming and restricting an idea is to make it public as soon as possible - and acknowledge all influences..... Luke Atkinson's NPA rake system could be one possible example of this, but the following example is able to use part of that because he is not into "owning" good ideas. Good ol' Luke and NPA.

Some years back I was sitting in Florida with Bill Patterson trying to release my hand from his steel desk after he pinned it there with some really powerful small magnets..... These Southern boys have a crushing sense of humour.

Some nine months ago, Richard Smith of Smith Engineering, Cumbria, sent me a simply switchable (on&off) magnet he had devised as a prototype handtool - but the design was too complex and expensive and the magnetism too weak (got the oldest deminer I've ever met to try it in Iraq - and he was polite when it fell apart: sorry Richard).

Three months ago I was watching deminers in Sri Lanka exposing mines with a Brush-rake - safely. And I freely admit that I was one of the cynics before I saw the complete system, and before I tried to set off a T72 with a Brush-rake - without success.

And two months ago I was watching deminers in Cambodia sifting soil over a speaker-magnet in the search for elusive fragments that slowed down their advance to a snail's pace..... Most of the spoil missed the magnet and they were reduced to rubbing the magnet on the ground to try to catch fragments. I have seen that before - with pieces of weak and broken magnet used optimistically by many groups, but it was CMAC who had made the best use of it.

Using the ideas of Bill Patterson, Luke Atkinson, CMAC, and no doubt others I have brain-picked over the years, I put together some simple magnetic attachments for existing tools.

Just a couple of weeks ago I watched deminers using magnetic Brush-rakes on heavily fragmented areas and finding those within 2cm of the surface with the rake - so speeding up metal-detector clearance dramatically. They also exposed the tops of shallow mines. In an identical area, deminers using clip-on trowel magnets had the same de-fragmentation success, and were able to sift soil (as they excavated detector signals) efficiently because the magnet extended along one edge of the tool (both sides). [First they passed the back of the tool above the ground surface and the powerful little suckers dragged frag out so that it clacked on the trowels (trowels were blast-resistant stainless-steel and unmagnetised). If that didn't work, they scraped the ground surface flat with a lateral movement (no downward pressure) and rolled the spoil over the magnet clip.] When both failed, they prodded and dug and tipped the spoil over the magnet as they did so. All ferrous fragments were found with the magnet.

These small neodymium magnets are far more powerful than anyone expects - and very light. A small and inexpensive addition that will make clearance faster almost everywhere - and very much faster in heavily fragmented areas. The Brush-rake version has a magnet that is moved from rake to rake as the rakes wear out. The magnetic tool-clip fits any sharp-edged tool (recommended for use on stainless-steel tools only), and changes from side to side for the right and left handed.

Incidentally, I collected the scrap from three minefields in Southern Mozambique and sorted the ferrous from non-ferrous. The ratio was more than 50:1. OK, I know that's only one province, in one country, at one time, but it's the best I can do so far.

So, I took the magnets to two commercial demining suppliers and gave them away. Their contact details are available on request. As always, I retain no commercial interest.

I cannot guarantee just how "safe" the magnetic Brush rake really is to use on AP blast mines - my first hand experience of failure to initiate is limited to two types (and reliable hearsay to three more) but it should be "proven" that the sprung-tines are acceptably safe with all common types (in predictable condition). And the design should be optimised for efficiency, safety, durability, etc..... [Yes, "in predictable condition" because no excavation system is safe when the mines are not in that state and because the Brush-rake allows the user to be standing a metre and a half back in visor and armour, so is inherently safer than most tools.]

Optimising rake design and use is part of something I would like to see happen for Atkinson rake system in general - and I apologise to Luke for speaking out of turn in this manner. In brief, the Atkinson/NPA system uses rakes for excavation and detection - without metal detectors - and is both effective and impressive. It gives a higher confidence of full clearance to depth than any other excavation system I have seen (which is all common manual methods, plus a few uncommon ones). But that system is another topic.

So, I have been asked to say something about the magnets now to ensure that no one pops up and claims commercial "ownership" of something too simple to be owned. I learned about the Brush-rake from Luke and have simply adapted it for use with metal-detector systems by adding magnets. The magnetic-clip for handtools is an obvious incremental advance on what is already done in some places.

The magnetic Brush-rake is probably the only real tool advance since blast-resistance became an achievable issue - and the only tool advance there has ever been with such a potential for safely increasing speed of clearance with metal-detectors. Lightly removing cut vegetation, leafmould, and loose topsoil as well as fragmentation - with the added advantage of exposing the tops of shallow blast mines.


Subject: Re: rakes

Hi Andy

Thanks for this, can you post some images... you know I don't like reading, as you know we intend to deploy towed magnets behind the Armtrac and we'll see what we get.


Subject: Re: magnets v batteries

Thanks for the positive feedback on the magnets and rake message. On and off line, there has been no real disagreement. Yes, anywhere with tilt-sensitive fuzes and/or tripwires it would be "unwise" to use a rake - without doing a tripwire drill and cutting the vegetation first. Which is always true for any method used in those areas. When tilt sensitive fuzes such as are on the PROM-1 and V69/J69 are in dense undergrowth, I believe that the only safe way to approach is to cut the undergrowth with a protected machine (which is where mini-flails should come in). The accident data from the Balkans confirms that, with numbers of deminer victims dropping away dramatically as machines began to be widely used.

Within suitable constraints (as for everything), I think that the magnet brush-rake would be safe to use in most places - as long as the inability to detonate AP blast mines with it can be confirmed. Part of the reason I am uncertain about that is the fact that the brush-rake is sourced from many manufacturers, each using slightly different methods and materials (sometimes from batch to batch). The basic gardening tool is then adjusted for the purpose - which always means fitting a stronger handle socket, and also means angling that socket appropriately when adding a magnet. Without some testing of materials and designs, I cannot be confident that the spring in the tines will always prevent a concentration of pressure - or that the tool will stay together in an accident. Meantime, whatever AP blast mines you are finding, I suggest that you use the rake to try letting one off (with only a det in it, of course) before deciding how best to use it.

This is available now from one manufacturer.

The strip of magnets is held in a stainless steel clip and stands away from the rake with rubber spacing. That and the angle of the head allows the magnet to be close to the ground without the "heel" of the head striking the ground.
The magnet clip for a handtool is also very simple - and fits any sharp edged tool (stainless is recommended to keep the magnetism in one area).

With the clip in place, magnets are on both sides of the blade - so letting you pass the back over the ground without quite touching - and to tip spoil over the magnet on the inside to find fragments as you work.

As far as Luke's full Rake Excavation&Detection System (REDS) is concerned, I am very impressed by its speed and thoroughness in soft to medium ground. But I have also seen it used on really hard ground with lots of large roots and rocks. In those conditions, it probably has little advantage over other excavation methods. But in those conditions, the work required to bury mines deeply means that usually they were not - so a magnet-rake and a metal-detector would probably be the way to go.

Usually, the only reason for excavating an entire area is that metal-detector use is stuffed by too many fragments and/or electromagnetic ground. Ground compensation can usually handle the latter - and the magnet-rake could be the answer to the former. BUT, the REDS system has another advanage entirely. It is very cheap indeed - and if you do not have to buy a detector or batteries at all, that can be very good news - especially because it is so thorough that you could hardly miss a coat-button, never mind a mine.... I would like to see just how far it could be improved by refining the Harrow rake-heads for tougher ground. It could be the answer to indigenising a lot of programmes (making them work within the local economy), offering a real alternative to draining demining money into Duracell.


Subject: Re: magnets v batteries

At the risk exposing myself as a little slow ........ what are the magnets attached to the rake for?

In the Sri Lankan scenario there are some 8 or 900 deminers out there armed to the teeth with rakes. The main advantage of the rakes is that the entire tool kit cost $25 per person, as opposed to somewhere around $2,500 for a single detector with added excavation tools also required.

By adding magnets are you suggesting a metal free end product. I don't see the value added unless you use detectors for QC. Don't you just get magnets clogged with metallic debris and dirt if you attach them to rakes.

Or maybe I'm missing something. Those were magnets attached to either side of the rake weren't they??

Sorry if I'm not picking up on a fundamental here but people have asked me and I am at a loss.


Subject: ARES about REDS

Sorry if I confused the issue.

Of course, you are right. With the Sri-Lanka Rake Excavation and Detection System (REDS) there would be no point at all in using magnets. The system is complete, although I think there would be many advantages in refining the tools to ensure consistency of quality and enhance performance (on hard ground and against roots). I call it REDS because the word "rake" still makes many people wince. As you know, these rakes are not your conventional snake-in-the-grass garden rake that blacks your eye when you step on it. The alternative name was the Atkinson Rake Excavation System - but the resultant acronym seemed a tad unsuitable.

Quite apart from REDS, the Brush rake with a magnet can be used in metal-detector systems to speed up clearance dramatically. In Sri Lanka, if DDG or RONCO and the army were to combine magnet-rakes with their metal-detectors, there is evidence that they would find really dramatic increases in speed in any area of high fragmentation. The inability to detonate AP blast mines with a Brush-rake (that Luke discovered and which needs confirmation against a wider range of mines) means that the use of the tool outside the REDS system can be seriously considered - and I have seen it working well.

And of course you are also right that the advantage of NOT spending money on detectors and batteries is obvious - and rather attractive - which is why refining both rakes in the REDS system to extend their efficiency in varied ground conditions is attractive. Problems with quality control in rake manufacture are also frustrating - and could easily be addressed.

The magnetic brush-rake is for use with metal-detectors - and is already out there for purchase.

The REDS system is an alternative to metal-detectors, and is also available now-now. I believe it would probably benefit from some refinement of the tools to extend the areas they can be used, but I think that  the NPA REDS system itself (including internal QA) is more-or-less splendid as is.

Hope that is clearer than my previous.


Subject: clearly still unclear..

In most places, metal-detectors and the human eye are the main means of detection. Where dogs are used, men generally follow with detectors. Where machines are used, men often follow with detectors. So metal-detectors are still the most common tool used in mine/device detection. Trouble is, they usually detect much more than what we are looking for.

Something between 300 and 500 fragmentation items to each mine/device is common (only a few places really count, but based on available figures 300:1 is low). Most of these items are not deep Sean - although no one records that, so that statement is only based on my own experience. The commonest fragments (found over a wide range of countries) are pieces of mine/device fragmentation, nails or parts of nails, bullets, bits of mines/devices/fuzes, bits of barbed wire, bits of tripwire, bottle tops, nuts&bolts, flakes of rusty cans, and flakes of unidendifiable trash (not listed in order of frequency). Current investigation implies that the ratio of magnetic to non-magnetic fragments may be in the region of 50:1.

Signal excavation drills usually require each fragment to be investigated as if it were a mine - withdrawing around 20cm, then digging and prodding to advance at the clearance depth towards the detector reading. Most groups encourage the deminer to eyeball the area and pick up visible scrap first. Many of the scrap items are well camouflaged by colours ranging from black to red (some deminers believe that fragments take on the soil colour, but I have not checked whether that has any truth in it: strangely, it does seem to). A few groups get the deminers to pass bits of magnet over the area - and they find some fragments despite the weakness of their magnets and the fact that they frequently have to rub the ground with the magnet to attract anything at all.

The magnetic Brush-rake is used after getting one or more metal-detector readings (so also after vegetation cutting, tripwire drills and any other area preparation). The strength of the magnets is high. The light raking of the ground surface loosens fragments. The magnet does not touch the ground, but must be close to it. If close, the fragments are attracted to it with enough force to frequently "clack" against it. Small flakes of rust, metal dust and bits of tripwire will jump onto it silently.

The magnet-strip is positioned near the rear of the rake-head. This means that it does not alter the spring in each of the tines. The flexibility of the tines is a safety issue. If the tines were stiffened in any way, they might be more able to transmit pressure onto the mines. To get the rear of the rake-head close to the ground while still allowing the user to work in a standing position, the rake-head must be angled to the handle. It is desirable for the user to be standing for comfort, and to maintain distance from any AP blast initiation. So the rake-head is angled. Only the tine-tips touch the ground, and they need not be any closer to the user than they are without an angled rake-head (depends on where the two metre handle is held).

The deminers who have used it gave it an almost unconditional thumbs-up. The "almost" comes in because a couple were concerned about the possibility of detonating unfamiliar mines with it. Their concern is very reasonable and should be checked out.

The complete magnetic Brush-rake as it stands is simple and cheaper than three sets of Duracells for a Minelab. Most of the cost is the magnet, which is moved from head to head as the rakes wear out. I believe that the whole thing could be improved (in terms of safety, durability and consistency) without making it much more expensive.

So the magnetic brush-rake is intended only as an additional tool for use in metal-detector clearance systems.

The brush-rake without a magnet has been used on-and-off since Gulf War 1 - and probably before. It was used in PMA-2 and PMA-3 areas of Kosovo - to remove surface leaf-litter in forested areas. When used for this, it happened to expose the mines without any detonations. I would like to check whether that was luck - or whether it really is acceptable to expose a PMA-3 with it. The PMA-3 pressure plate is the entire top part of the mine and the friction fuze seems likely to be particularly sensitive to offset pressure, even when that pressure is very light. Even if detonation is likely, it may be possible to design rake-tines with enough flexibility to reduce the chance of that.

It can be argued that, when the user is suitably protected, a small risk of harmless initiation is an acceptable return for a significant increase in speed. Given that there are unintended initiations with all manual methods and most of them put the deminer so close that severely disabling injury can occur, I think that this argument is worth considering.

How significant would the increase in speed be? Depends on the level of fragmentation of course - and whether there are any deeper signals to be investigated. Current evidence suggests that, if all the fragments were ferrous and in the top couple of centimetres, and there are no deeper readings, more than ten times faster than metal-detector based clearance done without a magnet.

[Noel: The handtool magnet strip shown with the picture of the trowel is actually two magnet-strips folded inside thin stainless steel. The magnets are "back-to-back". The clip slides over the sharp edge of any tool, putting a magnet strip on both sides of the tool blade. The magnets are strong enough to grip each other through the tool blade, and the clip stays in place without being "sprung". Having magnets on both sides allows the back of the tool to be passed over the ground without touching it - and also allows spoil to be tipped over the blade, catching fragments as they pass. It is really very effective and has scored an entirely unconditional hit with the deminers who have used it. I'll show you it.]

[Sean: In the REDS system, it is not necessary to make the heavy Harrow-rake even heavier for use on hard ground. It could be significantly improved by designing the points for penetration (as opposed to scratching) as the rake is drawn forward, designing the inside edge of the tine curves to cut, trying other tine curves to increase cut and also increase the "lift" of a wider range of encountered items, etc, etc. It works on hard ground even now, (in Somaliland as well as Sri Lanka) but I think everyone agrees that the tools could do with being refined/optimised.]


Getting hold of magnets

To purchase magnet rakes and/or clips, contact:

Security Devices Pvt - and speak to Oddia.
Telephone: (+263 4) 487064 / 5: Fax: (+263 4) 486885
Email: oddia(at)

I handed them over and walked away, so please let me know how it goes......

I have recently bought neodymium bar magnets from:


Paul Schumacher adds: "Here's a trick I learned from a prospector in AZ back in '72. Put the magnets used to pick up ferrous scrap in a plastic bag or sheet. This makes it a lot easier to remove magnetic material off of the magnet by inverting the bag and pulling loose the material."

If the magnet is very strong (Neodymium or Samarium magnets), and if the mine has much ferrous material in it, it could be lifted by the magnet or pull the magnet down onto it.