Funding constraints imposed by the cost of COVID aside, the Good, the Bad, the Damned and the Ugly have just advertised for a software engineer to design and write an entirely new accident database for Mine Action. (Well, the Damned have put out a job advert.) Presumably some ill-informed donor has been persuaded to part with some money. Ill-informed because using a software engineer to design a database they will never use is exactly what GICHD did with their RAPID spreadsheet fiasco. But there is a difference this time because the Good, the Bad and the Ugly (the three biggest INGOs in HMA who are referred to as ‘the Trinity’ in Geneva) will decide what goes in and what is shared. So that must be all right... surely?
In terms of past honesty and transparency about accidents, the three include the best (NPA), but also the worst (HALO). I suspect that the Damned (GICHD) has not told the donor that it has done this before – TWICE – and failed to gather any useful data (or share it). I very much doubt that they have told the donor that a useful bit of software already exists – one that deliberately uses the Microsoft Access database engine so that sharing it and its updates should be low-cost and easy. Instead the Damned have decided that an entirely new and bespoke database engine must be made so that use of the software can be restricted and that they will have complete control of what is recorded, its maintenance and its use. Being secure and restricted sounds reasonable? Not when we know how they have done things in the past.
First there was my (always anonymous) Database of Demining Accidents and Incidents started with help from Col Z and US ARMY CECOM NVESD (as was) – which provided evidence that was so useful when drafting the original IMAS that both UNMAS and GICHD wanted on board. They gave letters in support of my data gathering and arranged some UK money to improve the software (I knew what was needed, but could not afford it) and so the GICHD Database of Demining AccidentS was born.
Yes, the ‘DDAS’ is not a real acronym, but the GiHAD has never been good at those. GICHD was supposed to gather accident data for me to add while I designed and managed the software rewrite. I was obliged by contract NOT to gather reports myself. My ‘flat’ database had to become several databases inside each other to simplify both data entry and searching (this is known as a ‘relational’ database). At my insistence, the database was distributed freely on CD. Data from it was never used to embarrass the demining agencies because I had designed it to avoid that possibility. It was requested by INGOs and widely distributed amongst field people at the time. I bug-fixed several software iterations and got it working well – but GICHD failed to gather a single accident report over more than two years – during which there had been many accidents.
Their failure to gather accident data embarrassed those who were nominally managing me (it was all a bit ‘Ugly’), so they made no objection when I started gathering the data again myself (still with letters of support). The IPR for the software was always shared by me and the developer, and could be used freely. The software engineer offered to sell GICHD the coding (which was a simple front-end to the latest Microsoft ACCESS database) but they declined. I continued to gather data and use the database to inform IMAS revisions for another five years – always answering questions in detail (with the evidence) and making the data freely available to the industry. I even put a much used version on-line. It was successful and influential, so someone at GICHD decided that it was time they took it back. With no money, they used an in-house ‘talent’ to make a spreadsheet with nonsensical drop-down choices that would reduce a 4000 word accident report to a few inaccurate phrases. The thinking seems to have been to make it RAPID and very easy for MACs to report an accident, so meaning that they would do all the work for GICHD. They could give its upkeep to an intern and claim ‘job done’. Rubbish in, rubbish out, of course. And soon the MACs stopped sending these silly summaries because they could not find a box to tick (and got nothing back) so GICHD’s approach was a complete failure.
When announcing the release of their RAPID spreadsheet, UNMAS and GICHD told MACs to stop sending me their accident reports – and most obeyed. So now, no one was gathering the real data. I kept going – albeit erratically because I had to earn a living – and kept asking GICHD to take over the real database. They kept on asking me for data from it, but had no interest in taking it over. Two years ago James Madison University (JMU) renamed it the Accident and Incident Database (AID) and hosted it in their data repository – well done JMU. It has been extensively used over the last ten years (including by GICHD). Even this year, I have been to give working copies containing their data to Lebanon and Kosovo (to allow the ‘evidence based’ risk management that is required in the IMAS).
But now it seems that some poor donor has been duped into starting the whole process again. With years of development delay, they will let those organisations with most to conceal decide what data is gathered and what lessons are learned – all supervised by a bureaucracy that has repeatedly proven itself incompetent in gathering and using this data. Yes, I know that GICHD has become a finishing school for ex-Ugly Saints who, by definition, always start by knowing all they need to know. True, I can also come over as arrogance personified myself, but I cannot help observing that favouring semi-articulate, half educated macho arrogance (male or female) has always been Geneva’s problem. Yes of course that’s a problem in UNMAS too and Agnes was a disappointment to those who thought she might bring real improvements. I am told that she has resigned this week – so maybe her replacement will bring a breath of fresh air? (It’s the ‘ban bunny ‘in me that cannot help having these little surges of illogical optimism.)
Hey, I know this is partly personal. I have dared to criticise the self acclaimed UNMAS and GICHD experts and they cannot answer me, so have declared me persona non grata. Do I obstruct the path to their food trough perhaps?. Fine. The real accident database is now with JMU. Support it. Give them data. Let really independent people assess and analyse it to produce coldly self-critical lessons learned. Improve the gathering and analysis systems constantly – and let the industry start to manage risk based on the real evidence. Job done. But please, no more jobs for the boys in the mountains who have had their chance at least twice. Send accident reports to firstname.lastname@example.org.
You thought I had retired? Yeah well. Anyone wanting to talk – go to the Croatia Symposium thing in September. Anyone wanting to thump me – er, why not try using words instead? It really is time that UNMAS/GICHD grew up, stopped pretending that the team must be defended at all costs, encouraged self-criticism, and tried to answer me with the honesty and transparency of the Quality Management approach to continuous improvement that is (ironically) required in the IMAS - and which they have the hypocrisy to impose on national MACs and NGOs.