Original URL: https://www.theregister.co.uk/2007/11/11/stockwell_one_systems_failures/

The broken terror systems that killed de Menezes

A study in systemic failure

By John Lettice

Posted in Policy, 11th November 2007 11:02 GMT

Analysis Killers were on the loose and, to quote Friday's Guardian, "terrible risks had to be balanced... it flowed from this that something might go terribly wrong." Except, as the Stockwell One report into the de Menezes shooting makes clear, that's not exactly what it flowed from, and given the systems in place on the day, "might" severely understates the position.

The report quite clearly gives the lie to Met Commissioner Sir Ian Blair's claim that the incident was an "isolated breach", and that there is "no evidence at all of systematic failure by the Metropolitan Police." On the contrary, the systems of command, control and communications used were as a matter of design virtually pre-programmed to fail, to the extent that the life of Jean Charles de Menezes was in severe peril from the moment he stepped into the street. His death was not an 'unfortunate accident' but a direct consequence of systemic failure of the Metropolitan Police systems and procedures deployed (in this instance) for the surveillance and interception of suspected suicide bombers.

It would be accurate for police to complain that the death of Jean Charles de Menezes was 'beyond our control', but how did this, and effectively the entire operation, get 'beyond our control'? The answer lies in the intersection of a number of systems and procedures, faulty ones, several of them virtually pre-programmed to fail. Here we examine how that happened.


The inability of police Airwave radios to operate on the underground will tend to be seized on as the operation's main communications problem, but actually it's of little consequence; practically all that was going to go wrong had already gone wrong by the time de Menezes passed through the ticket barriers, and when the firearms officers (CO19) went through, his fate was already sealed. From the logs, it appears that little more than a minute elapsed between the firearms team going underground and de Menezes' death.

Officers involved in the operation on the ground used two radio systems primarily, with Cougar radios for the surveillance team, and Airwave as the general issue communications system. The CO19 firearms team had Airwave but also had the ability to monitor the Cougar traffic, and mobile phones were also used (so Airwave's working well, then).

The most obvious problems don't stem directly from lack of interoperability and/or inadequacy of the communications systems, but from the way communications were structured. Information from the surveillance officers on the ground was - at least in theory - collated by a surveillance monitor referred to as "Pat" in the report, then passed on to the control room and Commander Cressida Dick. The investigation found no evidence of any of the surveillance officers positively identifying de Menezes as "Nettletip" (the police codename for Hussain Osman), yet officers in the control room claim that they had been given positive ID.

Dick says that Pat told her that surveillance believed it was Nettletip, while Pat claims no recollection of this. CO19 officers claim to recall hearing a surveillance transmission saying it was "definitely our man", while a surveillance log altered after the event said either "definitely our man" or definitely not our man". The latter may have been altered to correct an error in the log of a sighting by surveillance officer "Laurence", who had reported that it was not Nettletip.

Says the report: "Despite the belief within Room 1600 that Nettletip had been identified it should be noted that every entry on the Surveillance Running Log refers to the person as being 'U/I male', meaning unidentified."

The report notes that the ground surveillance team all appeared to believe that they did not have a positive sighting, while the control room believed that they did. It looks very much as if faulty filtering and Chinese whispers effectively manufactured this situation, helped along by individual officer's fears of the consequences of making a mistake. This would tend to make them extremely cautious of saying that it definitely wasn't Nettletip, meaning the possibility that it was him was pretty well embedded in the system. These doubts would be filtered upwards, and the same fears would lead superior officers to give undue weight to a single claim of 'might be' over half a dozen of 'probably isn't'. The claims of CO19 that they heard "definitely" illustrate that direct monitoring in the control room isn't the answer either - nobody admits to saying that, and they may simply have missed the word "not".

Command and Control

The communications problems inherent in the way the Met was running multiple teams from a remote control room suggest a system that isn't actually fit for purpose as regards tracking and intercepting a suspect, rather than one that just didn't work well on the day. But the command and control systems come out badly too, sometimes for related reasons, but not always.

The report notes that Commander Dick viewed it as natural that she, as the senior officer, should be in charge of the operation, and that is no doubt how it works as a matter of course at the Met. But she was presiding over a system that was - as we've seen - incapable of passing timely, accurate and undistorted information to the control room, and which also failed in the other direction, over the interpretation of her 'stop' command to CO19, who claim to have thought it was a shoot to kill order. The initial deployment plan called for surveillance, CO12, to monitor those coming out of the building and for CO19 to detain them at a safe distance. CO19 wasn't however in place (as detailed by Lewis Page), so neither de Menezes nor any of the other people coming out of the building were stopped and questioned according to plan.

More autonomy on the ground, or better, command authority there, would quite possibly have resulted in a contingency plan kicking in. The other people leaving the building were presumably left alone because they were identified as definitely not Nettletip, and de Menezes would probably have been stopped and questioned (stop and question was the plan), rather than shot a short time later.

De Menezes was neither stopped nor ruled out as Nettletip because initially the officer watching the exit was unable to get a clear look at him (because he was having a pee). Arguably de Menezes' chances might have been better if at this juncture he'd been misidentified as Nettletip. Would they have let an identified terror suspect get on a bus, or would CO12 have intervened? It's a possibility - but he boarded the bus, his identity shrouded by doubt. The capability to make decisions on the ground was also decreased because the ground commander was with CO19, not in the vicinity of the monitored block of flats.

A word here about CO12's and CO19's roles. The CO12 officers were armed, but were only intended to stop a suspect in the last resort, on the basis that they did not have specialist training, while CO19 officers did. This differentiation wasn't specific to this operation, and can perhaps be seen as a further sign of institutionalised risk aversion in the police.

Because in the ideal scenario CO19 would do the stop, and because no word came from on high telling CO12 to do the stop, de Menezes was allowed to board two buses. Faulty communications made it difficult for Commander Dick to judge how close CO19 was, and therefore how feasible this plan still was, and inflexibility did the rest. Risk aversion right at the top? A contributing factor.

Intelligence and Risk

Jean Charles de Menezes was wearing jeans, a t-shirt and a light denim jacket. He was not carrying a bag. To the lay observer he was quite obviously not wearing or carrying a bomb, and this should have been apparent to the surveillance team, who observed him closely over 30 minutes. But there seems to be no record of communications indicating that anybody was actually interested in what he was wearing, and when CO19 finally saw him in the last minute, they were convinced that he was a bomber ready to detonate. So how could that happen?

The Register has argued in the past, and will continue to argue, that the claims made by the security services regarding the danger of the 21st July bombs have been severely inflated. They were faultily constructed from doubtful material, and the detonators failed, four times, to set off the main charge. We are aware that the prosecution in the bombers' trial took a different view, but we are also aware that the expert witness put forward by the defence disputed this view. We should also note that the bombs were bulky, packed in rucksacks.

At a pre-operation briefing on the morning of the 22 July Commander Dick "was left in no doubt that the events on the 21 had been serious attempts and that detonation could have been achieved by simply putting two pieces of wire together." This would, presumably, be if putting the wires together had set off a powerful enough detonation to ignite the main charge, meaning that the 22 July 'bomber' would have a more effective detonator, a more effective main charge, or both. Is that likely?

Go just a little further down the risk assessment road and you wonder about the probability of a surprised suicide bomber with, QED, no pre-arranged escape plan, doing anything as implacable as building a bomb version 2.0, as opposed to running like a rabbit (which is what they all turned out to have done). And he wasn't wearing a rucksack, so any bomb would have to be a concealed suicide belt or harness - how likely is that? And if they did have these, and they worked, why didn't they use them on 21 July? Who would have used these 'spare' explosives if they had all blown themselves up?

Some of that's certainly possible, but even if you're anticipating worst possible case (risk aversion, again), the overall risk seems to take it into utterly improbable possible case. Attire should be factored into the risk assessment, unless you've already decided to anticipate the worst possible case where a devastating bomb that you are entirely unable to detect is present. The briefing to CO19 has more to say on explosives and suicide harnesses.

Detective Chief Inspector C from the anti-terrorist branch, CO13, was the "Silver Commander", effectively the ground commander for the operation. He based himself with CO19 and CO13 at the Nightingale Road Police station initially, and according to the report joined them in playing "catch up" throughout the operation. He briefed CO19, giving them information about the explosives used on 7 July, confirmed links between the 7 July and 21 July bombers, and "confirmed that the terrorists had the capability to attach a device to themselves that would be difficult to detect..." They were "deadly and determined and 'up for it.'" Earlier, CO19 had been briefed that they might have to use unusual tactics, and might be asked to do something they had never done before.

Some months prior to the July bombings the Met itself had been briefed by Israeli security on suicide bombers, so although the UK has yet to see terrorists capable of constructing effective concealed suicide harnesses, the Met was receptive to the notion that they might be out there already.

As the report notes, "The briefings that the officers received could only have heightened their desire to arrest the terrorists and add to the apprehension concerning the danger to which they were being exposed." CO19 were therefore expecting to confront a dangerous terrorist with a concealed bomb - wearing a light jacket and t-shirt wouldn't necessarily be enough to save you.

The last minutes

As de Menezes approached Stockwell station, Commander Dick asked for the surveillance team to give a percentage indication of how certain they were. Another officer asked for a score on a scale of one to ten, and this was described by one surveillance officer as a ridiculous question, impossible to answer. This week the Independent reported that the Met has now implemented a new system of classification of ID for surveillance teams, with three levels of certainty. So a score of one to three.

If Commander Dick was told that it was "definitely" Nettletip, she was told at this time. Similarly if CO19 heard "definitely", they heard it as they were moving in on Stockwell station. They would have been expecting a confirmed bomb suspect who wouldn't necessarily show any sign of carrying a bomb, and they were ready to employ "unusual" tactics if they were asked. De Menezes' chances were now slim, and rested on CO12 stopping him before CO19 got there.

As de Menezes approached Stockwell, Commander Dick was insisting that CO19 stop him - "don't want your people going up to him", she told SO12. But, "can't let him down the tube". CO19 weren't there yet.

At 10.03 de Menezes entered Stockwell station, and at 10.04 Commander Dick ordered: "must be challenged before going down the tube. Stockwell tube." But "No stop without 19". At 10.04 again, "Stop him". 10.05 "12 to do it." But also in the log for 10.05, "State red. SO19 doing stop do not let surveillance interfere."

"State red" means that SO19 has arrived and assumed primacy, and the operation is beyond the point of no return. You know what happens next. ®

* From very early on in the investigation process, 'sources' sought to undermine Jean Charles de Menezes. This process might be said to have culminated in the closing speech of Ronald Thwaites QC, for the Metropolitan Police defence in the recent trial for health and safety offences. The IPCC report does not support Thwaites' picture of a twitchy illegal immigrant confused by drugs and who "reacted precisely as [police] been briefed a suicide bomber might react." Nor does it support similar claims made in the evidence of Met officers, including that of Commander Dick.

The Register noted some time ago that a Home Office statement on the subject of de Menezes' immigration status (which was questioned by 'sources' immediately after the shooting) did not confirm categorically that the Indefinite Leave to Remain stamp in his passport was forged. We would therefore like to draw readers' attention to the note on page 21 of the Stockwell One Report which states: "Evidence emerged during the course of the criminal trial into the Health and Safety charge that Mr de Menezes was lawfully in the country on 22 July 2005."