Open Source Humanitarian Demining Project

[Humanise]

I'm trying to start a new Open Source project for the design of a metal detector suitable for Humanitarian Demining. I have started a website for this project at << URL deleted by berkeman -- please Google this person's account name to find more info on this project >>. If you agree with the objectives can you please promote/advertise the project. Better still, why not register and take part in the discussions/design.

There is small amount of Electronic design required. Most of the project will be done in software but an understanding of electronics would be a big help to anyone taking part.

Objectives of this project are as follows:

1. Leverage the power of Web Based Open Source Collaborative Projects,
2. Utilisation of New Technologies such as recent 24 bit sound cards,
3. Very Modular Design with Well Defined Interfaces between Modules,
4. Documentation of the Technology and
5. Support of other Beneficial Uses.

1) Leverage the power of Web Based Open Source Collaborative Projects.

A large number of important technology projects have been given Creative Commons or Commons like licenses such as the Gnu General Public License. Much of the development of these projects has been done by volunteers or through donations. Source code and documentation is normally freely available under the conditions of the license. These projects are generally referred to as 'Open Source'.

Web based communication is creating unprecedented opportunities for large scale knowledge aggregation. Web based social software tools and environments are enabling multitudes of people to connect, communicate and exchange information and knowledge in near real time and foster new ideas. This has broken down traditional barriers of distance and exploded the potential for 'collective intelligence' to take place.

2) Utilisation of New Technologies such as recent 24 bit sound cards.

One of the central reasons for the creation of this project was the recent availability of audio cards able to perform 24 bit Analog to Digital (A/D) and Digital to Analog (D/A) conversions at high speed on multiple channels at reasonable prices. The jump from 16 bit to 24 bit A/D and D/A technology gives quite spectacular advances in specifications. In the future it is likely that they will be available at even higher precisions, higher speeds and lower prices.

It is assumed that the initial projects will be based on Personal Computer (PC) compatible hardware and compatible peripherals with specially designed interface circuitry to interface the audio cards to coils. Both software created and software tools used for designing it will be Open Source software.

3) Very Modular Design with Well Defined Interfaces between Modules.

The design is created within a framework which splits the design into modules, with well defined interfaces between each. The project further splits these modules into tasks needed to complete each module or otherwise help to complete the design. This allows people or groups of people with skills in specific areas to perform a task or design a small module that will interwork with the remaining design. This should achieve the following:

* Bring each design task into being manageable for people who are donating their time.
* Allow people to utilise their own specific skills and knowledge.
* Greater scope for the project to achieve significant advancements in technology.
* Ultimately should give the project multiple versions of each module so that there is choice for those that utilise the results of the project. This should allow the creation of a number of different detectors to meet different needs such as hand held detectors, vehicle mounted multi coil visual indication detectors and detectors used in different environments.

4) Documentation of the Technology

This includes Documentation and Publication of the following:

* Previous technology,
* Current technology,
* Any technology advancements created by the project and
* Principles of any likely future advances.

This is needed not only to help the project, but to try to stop or reduce the technology being usurped by Intellectual Property (IP) ownership.

Related to the documentation objective, part of the project is to create a central Internet repository for the project's design, documentation, code etc along with the tools needed to organise and help the project such as a forum, wiki etc.

5) Support of Other Beneficial Uses

While the primary purpose of the designs is humanitarian demining, some designs or modules from the design are likely to find use elsewhere. Examples include:

* Anti terrorist and other security screening.
* Gold detectors and other geological exploration instruments.
* Archaeology exploration.
* Hobby coin detectors.
* Underground pipe/cable/leak detectors.
* Manufacturing uses such as detectors to help ensure contaminants don't enter our food supply.
* Detectors used for separation and recovery of different types of recyclable material in waste management.
* Medical uses.
* Research uses.

These other uses are allowed, encouraged and supported.


Please visit << URL deleted by berkeman -- please Google this person's account name to find more info on this project >>

for more details

 

[Astronuc]

Firstly, metal detectors already exist.

Secondly, landmines are already designed to minimize metal, or be minimially sensitive to metal detection.

Thirdly, mines or subterranean structures are imaged with Ground Penetrating Radar (GPR) which has matured over the last two decades.

 

[Proton Soup]

the need to use "24 bit sound cards" completely eludes me. what is the real purpose here? you don't need anything fancy like this.

 

[Humanise]

the need to use "24 bit sound cards" completely eludes me. what is the real purpose here? you don't need anything fancy like this.

To make a simple metal detector that can detect metal is very easy. The old Beat Frequency Oscilator (BFO) actually work and are very cheap to build but these would not be good enough for mine fields.

Trying to make a top quality one is very difficult and expensive and this is needed where you have mines with very little metal content (a fraction of a gram in the firing mechanism). Of all the possible ways to find mines (and there are a lot of them) the two methods that are most used are metal detectors and prodding with a stick.

Regardless of their heavy use, metal detectors are still not good enough. The number of false replies to correct finds is often at the ration of 1000 to 1. It makes if very expensive to demine and a lot of mine fields remain because of it.

The big challenge to metal detector design is to be able to design detectors that can characterize targets such that the false fines can be reduced. As one starts to understand the principles involved, there are a large range of target properties that can be examined but these will often require every drop of precision possible. If I could get even more advanced sound cards at a reasonable price then I would be using those instead.

Over the next few weeks I will be preparing a number of articles on the technology and the concepts of what could advance the design. I'm trying to write them in such a way that the reader doesn't require understanding too much maths or physics etc. I'll be placing them on the website soon.

Over the last few years, metal detector technology has advanced and I believe that it will continue to do so. Those in this project should help for it to do so.

 

[Humanise]

Firstly, metal detectors already exist.

Yes, but ones good enough for what is currently used for demining are still very expensive. Further, even though they are currently extensively used, they are still not good enough.

Regardless of this being an old technology, over the last few years there has been significant advances in their design

Secondly, landmines are already designed to minimize metal, or be minimially sensitive to metal detection.

This can be problematic depending on country/area but in general, better quality detectors are able to find these mines as they have a small amount of metal in the firing mechanism.
The people on the front lines using these machines are screaming not so much about this problem but about the problem that current metal detectors are very poor at being able to characterize the targets they find. The number of false replies to correct finds is often at the ration of 1000 to 1. It makes if very expensive to demine and a lot of mine fields remain because of it. I consider the big challenge to metal detector design is to be able to design detectors that can characterize targets such that the false fines can be reduced. As one starts to understand the principles involved, there are a large range of target properties and characteristics that can be used.

Thirdly, mines or subterranean structures are imaged with Ground Penetrating Radar (GPR) which has matured over the last two decades.

This is one of a number of promising technologies (along with dogs and rats) and this is one of the few that is actually being used in a number of areas. Regardless, the two methods that are still most used are metal detectors and prodding with a stick. Yes, regardless of all the technology, simply prodding every inch of a field with a wooden stick is still a method heavily used.

No single method will work for all areas (other than prodding with a stick). Metal detectors are likely to continue to be heavily used.

Over the next few weeks I will be preparing a number of articles on the technology and the concepts of what could advance the design. I'm trying to write them in such a way that the reader doesn't require understanding too much maths or physics etc. I'll be placing them on the website soon.

I believe that metal detector technology will advance further. Those in this project are able help for it do so.

 

[Proton Soup]

is the audio card not for audio, but a cheap function generator?

 

[mgb_phys]

is the audio card not for audio, but a cheap function generator?

A cheap high accuracy data acquisition device (ADC) with lots of DSP pattern-in-noise finding capability.

 

[Proton Soup]

well that would certainly be useful

 

[mgb_phys]

I've heard of a few robotics project using echo-location for 3D mapping the room just using the sound card.

Another big breakthrough is the graphics card on modern PCs - some have >500 CPU cores, which for a limited set of parallelisable operations outpower a supercomputer of a decade ago. It's being used for all the normal supercomputer stuff (http://en.wikipedia.org/wiki/Molecular_modeling_on_GPU) and things like real time wifi/cell phone encryption cracking!

 

[Proton Soup]

yeah, i was reading something recently about the proliferation of CPU cores being a security hazard, and that maybe they should be limited to some small number. but like you say, they're already used densely in GPUs. also, we've let the genie out of the bottle already, so some other nation will make them. we simply can't control the tech anymore. as for the cracking, i recently saw a screenshot of a guy that had gotten ahold of a bunch of hashes and was busily brute-force cracking the whole lot. that'll be one valuable database when he's done.

anywho, why don't deminers just plow minefields up with heavy equipment?

 

[mgb_phys]

anywho, why don't deminers just plow minefields up with heavy equipment?

Very expensive
Doesn't work for paddy fields / rocky ground.
Destroys farmland - all you have to do to make your home safe is plough up all the fertile soil

 

[MATLABdude]

Very expensive
Doesn't work for paddy fields / rocky ground.
Destroys farmland - all you have to do to make your home safe is plough up all the fertile soil

Whatever happened to mine flails? I remember WW2-era footage of these things whipping up a field and exploding mines as they went at it. Fairly cheap, low-tech, and pretty good at its job. I always thought that was the ultimate in demining, and that if you had a few of these, and a few years, you could demine most of the open terrain in a former war-zone.

I don't mean this as a dig at the original poster, or the goals of the project, but is open source really the term to use for this sort of thing?

EDIT: mgb_phys' post about paddies gave me a thought... What if you could flood an area with a few feet of water and then set off explosives (possibly shaped to go down and out rather than upwards)? The incompressibility of water would transmit the force better and generate an overpressure over a wide area (though I don't know how much water you'd need to make this true). It'd be like dynamite mine fishing.

 

[mgb_phys]

They do something similair to clear minefields in korea, a 20m long plastic tube filled with dynamite laid on top of water - light blue touch paper and retire to safe distance.

But it only works with mines <10cm below the suface (I suppose mud is a good insulator) - the problem is that if you have an entire country that could be mined it gets a bit expensive.

 

[Humanise]

Another big breakthrough is the graphics card on modern PCs - some have >500 CPU cores, which for a limited set of parallelisable operations outpower a supercomputer of a decade ago. It's being used for all the normal supercomputer stuff (http://en.wikipedia.org/wiki/Molecular_modeling_on_GPU) and things like real time wifi/cell phone encryption cracking!

I love this idea. I don't think that the humanise.org project needs it yet but if people come up with enough new ideas on how to analyze the data and how to make more use of more harmonics etc then it is always possible that it could be useful in the future.