Old technology hampers police in fight against moped gangs
POLICE are struggling to tackle the epidemic in violent moped robberies because they are equipped with outdated technology, The Daily Telegraph has discovered.
Knife-wielding thugs riding highpowered scooters have been carrying out scores of raids every day, often attacking victims before making off with valuable watches and mobile phones.
Last year in London there were more than 22,000 reported offences. Abdul Samad, a charity worker, was stabbed to death by two attackers described by a judge as “21st century highwaymen”.
But efforts to fight the scourge are being hampered by the lack of cuttingedge technology available to police officers on the front line.
Policing leaders say out-of-date smartphones and a failure to follow the hi-tech tracking tactics of the New York police leave officers a step behind the criminals they are trying to catch.
‘Officers have this modern technology at home yet when they arrive at work they take a step back in time’
‘Big data allowed us to target resources, to target key individuals. It’s not perfect but homicide has halved’
IT WAS about 10:30pm on October 16 in west London when Nathan Gilmaney, 18, and Troy Thomas, 17, climbed on to a moped from which they had removed the number plates and began an extraordinarily brutal four-hour rampage of theft, assault and murder.
Their first victim was in Maida Vale. Nazmul Chowdhury was stabbed in the back as he jumped on a bus. Within an hour, Ehimer Fierro-berbesi and his aunt Ana were robbed. After he had given them his cash, they stabbed him.
Almost immediately, Gilmaney and his accomplice turned their attention to Abdul Samad, a 28-year-old charity worker who handed over his phone. They drove off but doubled back when they realised they needed the phone’s PIN. Samad gave it and handed over his wallet too, only for Gilmaney to stab him in the chest, severing the right ventricle of his heart.
Samad crawled the few yards home, and died in the arms of his parents.
Yet Gilmaney and Thomas had only just begun. Minutes later they attacked Joseph Denby, who escaped, possibly because he was carrying an old phone not worth stealing. Juan Perez Buendia, who was attacked next, threw his bag at the feet of his assailants but was chased until he gave up his phone.
Having stopped to sell some of their stolen goods, the duo re-emerged around 1am to rob five more people and either punch or stab them, showing no mercy. At 2am, police gave chase. One of the fugitives removed his helmet, a common tactic criminals know can deter pursuit for fear of a fatal accident.
So grave were the attacks that the hunt continued anyway. The pair abandoned their scooter and ran. Finally one officer bravely tackled Gilmaney. A police helicopter equipped with thermal imaging cameras spotted his accomplice hiding under a nearby car.
In court they were found guilty of murder. It had been, prosecutors said, a spree motivated by “bloodlust”. But could it have been stopped earlier?
The answer, had it occurred in New York, is almost certainly yes. There, information from CCTV, number plate readers, chemical and radiation senaverage sors, 911 calls – even microphones that can “hear” gunshots – is fed to NYPD officers, all 35,000 of whom can access it through mobile smart devices.
Collectively, these technologies are known as the DAS – or Domain Awareness System. Built by Microsoft for the NYPD, it has been used for a decade, and in the past few years has added automated pattern recognition – of the kind that might have alerted Met Police last October that something horrific was unfolding in west London.
London has 10 times as many cameras, checking 38 million plates a day against a list of “vehicles of interest”. And while the NYPD has just 2,000 cameras scanning the streets (it can also tap into 4,000 private security cameras), there are an estimated half a million cameras in London.
During the Gilmaney-thomas trial, the Old Bailey heard that CCTV footage captured the attacks in chilling clarity.
“The police generate a huge amount of data,” says John Wright, whose 22year police career peaked as Britain’s leading officer in counter-terrorism technology, and who now works with the global tech security company, Unisys. “But I was just speaking to a senior Met officer this morning. His response was that it’s very, very difficult to interrogate that data.” The capacity for police to analyse CCTV video exists, Wright added, “but it tends to be used historically rather than proactively”.
He is not alone in questioning the use of data generated by modern policing to investigate crimes after they have occurred.
“Clearly DAS could have been very useful [in the Gilmaney-thomas case],” says Rory Geoghegan, who in the Metropolitan Police spent five years investigating gangs, and is now at the Centre for Social Justice think tank. “CCTV and number plate recognition – or indeed the absence of number plates – on moving vehicles could be analysed to trigger alerts in real-time.”
In contrast to NYPD officers’ access to real-time information, the Met’s progress has been “glacial”, he says.
Officers rely on a communications system called Airwave, whose data speed is 30,000 times slower than the home internet connection. Most officers on the beat have no direct access when away from the station. A few specialist anti-moped crime officers now have tablets with access to live information, but the majority use radio to check criminal records, sometimes waiting for half an hour for a response.
As long ago as 2013, the London Assembly was describing the situation as “incredible”. It noted in a report: “Officers have this modern technology at home yet when they arrive at work they take a step back in time.”
Four years later, the draft Police and Crime Plan for London 2017–2021 recognised that front-line officers lacked “mobile data tablets to enable them to work on the move”. Indeed, it was only in May that the Met signed a £44 million contract for an integrated IT system, CONNECT. Even if it works, it won’t be ready for several years.
Gilmaney and Thomas’s crimes featured in a spike of violence that saw knife crime in England and Wales rise by a fifth in 2017, to 39,000 offences – up from 26,000 five years ago.
According to ONS figures in April, gun crime, murder and robbery rose sharply, with reports of violence doubling in less than a decade, from 709,008 in 2008-9. Half that increase in knife crime occurred in the capital. Its murder rate went up 44 per cent in 2017 and the killing of young people, robbery and home burglary all increased by a third.
Even the police admit that violent crime is rising dramatically. “Data show that violent crime and the associated homicides, car crime and theft are all increasing significantly,” the police’s own statistician, Chief Constable Bill Skelly, concedes. “These are genuine and deeply concerning but the public should be assured we are doing everything we can to bring them down.” This usually means manpower.
In the past eight years, the number of officers in the 43 forces in England and Wales has fallen 20,000 to 120,000. For some it is tempting to pin the rise in violent crime exclusively on cuts.
Yet this is simplistic. Police numbers are returning to a historic norm after free-spending days under New Labour, when funding was increased by 40 per cent in real terms. Even then, a Labourdominated Home Affairs select committee cast doubt upon the link between cash and crime. “The reduction in overall crime levels does not seem to have been directly related to additional resources,” the committee emphasised in 2007. “We consider it unacceptable that the significant recent investment in the police is not being used to maximum effect.”
Today, it seems, the story remains the same. “Spending on IT has been huge,” says Geoghegan. “It’s not about money. It’s about innovation being stifled.” Innovation offers huge potential benefits to fighting violent crime. For example, analysis of so-called Big Data can help pinpoint where violence is likely to occur so officers can intervene.
Scotland is home to perhaps the most celebrated violent crime reduction story in modern Britain. Beginning in 2005, the Violence Reduction Unit (VRU) oversaw a new strategy based on treating violence like “an infectious disease”, according to its founder, Karyn Mccluskey, a qualified forensic psychologist with Strathclyde Police. She mapped the network of criminal gangs in Strathclyde, unearthing almost 200, and discovered that much violence went unreported.
Once she had established intelligence, the VRU adopted a three-step plan to “inoculate” those who had yet to be infected, and cure those who had. Step Two was a vow to help those who stepped away from gangs with jobs and education. Step Three was hearing personal stories from the mothers of those who died on the streets.
But the first step was clamping down. “We went through the roof with stop and search,” says Will Linden, acting director the VRU today. “We increased patrols because we needed – in medical terms – to stabilise the patient. We needed to stabilise communities and that meant more policing.”
Linden says the unit “invested heavily in mapping software – we recognised the geographic nature of crime. We were able to overlay it with health and education data. Analysts were creating something that police officers could understand and use. Big data allowed us to target resources, to target key individuals. It’s not perfect but homicide has halved.”
Then in 2012, the VRU established a homicide database for Scotland called Phoenix. Officers went back decades, populating it with videos, pictures, location and behavioural information from historic crimes. Immediately, patterns began to emerge. “There is a link between domestic violence and homicide, sadly,” says Linden. “Too many women die at the hands of their partners or ex-partners.”
Much domestic violence goes unreported, so police began to look for other ways of spotting it.
“Phoenix identified new opportunities,” says Linden. “Early on, for example, some vets asked us to look at cases of animal abuse. We recognised some names responsible for that violence. With Phoenix we began to see how animals were frequently used as coercive control over victims. Previously, making the link between animal abuse and domestic violence and homicide would have meant bringing together five or six different systems, spending months on research. It took us one afternoon.”
Police trained vets how to link animal abuse with domestic abuse and refer victims to support services. Now police have rolled out the scheme to dentists, beauty therapists and hairdressers who might spot signs.
South of the border, identifying patterns in the vast quantities of data is much harder. “There are probably 250 different police databases in England and Wales, from lost property to child protection,” says Alexander Babuta, a researcher and author of the report Big Data and Policing. Many police data sets are not compatible. Even when they are, analysis is almost entirely manual despite, Babuta notes, “software being available”.
The ramifications are clear for cases like the Gilmaney-thomas spree. “Police,” Babuta says, “are unable to take full advantage of the UK’S wide-reaching surveillance capabilities.”
“There needs to be a national operating model for data science and analytics in policing,” says Henry Rex, programme manager for Justice and Emergency Services at Techuk, which represents more than 950 UK firms.
Indeed, Babuta suggests technology demands may eventually force a rationalisation of the 43 forces in England and Wales, down to perhaps 10 or so, along the lines of the 10 Regional Organised Crime Units operating in England and Wales.
Nationwide, the Police ICT company set up in 2015 to address the issue has failed to consolidate the software used.
The benefits of predictive mapping, particularly for stretched forces, is clear. It has been shown to pinpoint crime with 10 times the accuracy of routine patrols, and at least twice the accuracy of intelligence-led patrols.
No wonder that most forces in the country are keen to use the tech tools at their disposal. “Can the police use data better?” asks Henry Rex. “The one word answer is ‘yes’.”
Smaller, nimbler forces like Durham are making strides, developing their own data analysis systems, but many officers find they often have to input the same details of a single case up to a dozen times. Many are also wary of algorithm-driven automated big data analysis. Some fear it will put them out of a job, or reduce the importance attached to human investigation. Will Linden cautions: “Big data analysis allows you to target resources better, but it doesn’t solve crimes for you.”
There are risks in careless use of analytics. But harmonising police data and allowing their mining in real time would allow officers to be more effective in crime prevention. This requires changes in access to data but also in police attitudes to data too. “Police decisions are driven by the risk of sharing [data],” says Henry Rex. “They must focus instead on the risks of not sharing.”