Paul Taylor was named the first national policing chief scientific adviser last year. He tells Beckie Smith how he’s working to improve the way science and technology is used in policing – and the “Del Boy moments” he’s had along the way
One minute, Paul Taylor is talking about the ocean, and the next, he is talking about dogs – neither of which CSW expected to cover in an hour-long video call with the national policing chief scientific adviser.
Taylor is illustrating his own surprise over the last few months at how many areas of policing are impacted by science and technology. In particular, he says, he was surprised by the breadth of scientific questions surrounding police dogs.
“They’re a really critical capability in policing,” he says. “You have issues around training and selection – because they’re very expensive, believe it or not. And some just aren’t fit for it, particularly the dogs that support frontline policing. They get put in some quite intimidating scenarios and they have to be brave enough to act, and act proportionately.
“Then there’s questions like: how effective are they at smelling various things, from explosives, to weapons, to multimedia, and so on? There’s loads of science around there.”
"Dogs are a really critical capability in policing. You have issues around training and selection – because they’re very expensive, believe it or not. And some just aren’t fit for it"
Another area he hadn’t expected to spend much time on was oceans. “I got schooled earlier in the week,” he says, recalling a meeting of the government-wide chief scientific adviser network, where he was asked if policing had a scientific interest in an ocean project the Department for Environment, Food and Rural Affairs was getting underway. Initially, he was inclined to say no. “Then my colleagues told me, ‘of course we do. You’ve got seaweed and algae and other things affecting the decomposition of bodies’.”
In the manner of a true academic, Taylor – a psychologist by training – lights up as he lists a few more areas he has delved into in the last few months. “Materials science is really critical for policing, because it goes into how comfortable uniforms are, it goes into body armour. It goes into how you’re building things, and so on. Behavioural science, that’s another thing that we’ve driven forward quite a lot over the last six months – because ultimately, policing deals with people, so we really ought to have a strong behavioural science capability to understand things like crowd intervention, how people are going to react when we come out of Covid, and so on. The list is endless.”
It is Taylor’s job to ensure the policing sector – which includes the College of Policing, the National Crime Agency, the Association of Police and Crime Commissioners, and individual forces – has access to evidence and advice on these and any other areas where science might contribute to crime prevention.
Taylor and his team are currently working on a strategy that will aim to show how policing can benefit from science and technology, and identify areas where policing bodies should concentrate their research and investment to expand the evidence base on promising technologies.
The strategy, Taylor says, will “articulate for the first time how a very complex landscape is going to interconnect”. It will complement strategies being developed by individual police capabilities – including the National Police Air Service, which has what Taylor calls a “very ambitious desire” to use drones to cut down on the use of helicopters. Doing that means addressing several “meaty, technological problems”, as well as getting the right standards in place.
Alongside the strategy, the National Police Chiefs’ Council is in the process of setting up a National Crime and Justice Laboratory which will, Taylor says, “allow us to do more strategic analysis of police and justice data to understand what’s going on in the system, and look for ways to improve delivery”. Creating this lab will meet a pledge made in the 2019 Conservative Party election manifesto. Policing bodies, the Home Office, the government funding body UK Research and Innovation and others will be collaborating on setting up the lab over the next 12 months.
Much of the work to address policing’s big science and tech challenges is done in collaboration with academics, and when Taylor was named as the first national policing CSA last spring, strengthening collaboration with external experts was one of the stated reasons for creating the role.
This is familiar ground for Taylor, who started his own career with a PhD on hostage negotiation before later going on to lead the Centre for Research and Evidence on Security Threats at Lancaster University – a hub for behavioural and social science for national security, funded by intelligence and security agencies and the Home Office.
He says there have been “a few Del Boy moments” where – needing an expert in a specialist area to weigh in on a complex problem – he’s excused himself from meetings to give an old colleague a call. “You’re halfway to a solution then – because you know the solution exists, it’s just not made its way to us [yet],” he explains.
In one initiative to strengthen collaboration with academia, the Economic and Social Research Council – part of UKRI – shared access to a catalogue of projects it had funded in the policing space over the last five years. That led to meetings discussing how researchers could work together, as well as a conference the police will run later this year.
In another effort to increase collaboration with external experts, Taylor has recently set up a police Science Advisory Council – a mechanism used by chief scientific advisers in government departments that acts as a “critical friend” to the organisation.
The police SAC includes university vice chancellors, chief technology officers of major tech companies and forensics specialists. It advises on strategic challenges and stress-tests plans, says Taylor: “It might be thinking about how we’re going to deal with digital forensics – what we should do to invest, how can we accelerate that work? It could be: how do we best engage academia and industry? How do we ensure that we have developed science and technology in a way that earns the public’s trust, using a variety of mechanisms to engage the public?”
One topic where the police may have a long way to go in winning public trust is in the use of facial-recognition technologies. A 2019 study by the Ada Lovelace Institute found public support for facial recognition was “conditional upon limitations and subject to appropriate safeguards”, and there has been vocal opposition to trials of the tech by different forces in public spaces.
In 2020, the Court of Appeal ruled that the use of automated facial recognition by South Wales Police over the previous few years had been unlawful – in part because the force had failed to provide evidence of an adequate data protection impact assessment. The scheme, which matched images of passers-by against people featured on criminal “watchlists”, also failed to meet its Public Sector Equality Duty as it failed to ensure the software had not embedded bias on racial or sex grounds.
Asked how scientists have responded to these concerns, Taylor says forces have “grown in their maturity” in recognising the issues that need to be addressed with facial-recognition tech. Studies on facial recognition software, for example, must use a sample of faces that meets particular diversity criteria. Committees at force and national level, as well as the Science Advisory Committee, meanwhile advise on and review protocols.
"Historically we haven’t been good enough at selling how much we invest in ensuring facial-recognition systems are not biased"
“Historically we haven’t been good enough at selling how much we invest in ensuring these systems are not biased and they’re not discriminatory. It really is at the top of the minds of those who are leading those programmes of work. We invest, we do trials,” Taylor says. He adds that the part of his role that entails engaging with the public on the use of such technologies involves “a little bit of myth busting”.
“I think sometimes people think we’re doing things with face recognition… that we’re not actually doing,” he says. He notes forces more commonly use retrospective facial recognition, such as analysing CCTV footage, than live facial recognition – although the Met and South Wales police forces say they use both. Last summer, then-information commissioner Elizabeth Denham said she was “deeply concerned” about the potential for live facial recognition to be used “inappropriately, excessively or even recklessly”, and that every investigation by her office into deployment of the technology to date had found illegality.
Another way the police can strengthen public trust, Taylor says, is through open science. “The general thrust that I’ve had since I’ve been in office is one of real transparency and really buying into the open-government and open-science agenda,” he says, adding that the Met and South Wales Police have published the findings of their facial recognition trials online.
Publishing information about trials, along with research findings and anonymised datasets, is “a really important way to build public trust so that the wider professional research community can interrogate what we’re doing and critique it and say, ‘actually, no, I think you’ve got it wrong or you’ve missed this’,” Taylor says. “I think that allows for wider public trust in the system.”
Later this year, the NPCC will launch a website – science.police.uk – that Taylor says will act as “a single gateway into science in policing” with news on what technologies are being used, alongside primers, a magazine and other information. “We have a federated model of policing for very good reason. But what that does mean is that sometimes, somebody in Merseyside doesn’t know what somebody in the Metropolitan Police is doing around, say, coding and data analytics,” he says.
As well as giving forces a better idea of what is happening elsewhere in the country, the website will act as a “signpost to academia and industry” of the NPCC’s areas of research interest. “If you talk to technology companies, they’re often a bit frustrated that they don’t know exactly what we want sometimes, and they’re not quite sure who to go to [to find out],” Taylor says.
“And it’s also a great way for other government departments to know what we’re doing as well. We’re very keen on cross-department join up and it’s a great way of allowing people to see what we’re doing.”
It is also being developed with the public in mind. The chief scientific adviser’s office funds around 50 research projects, and details of that work and its outcomes will be shared on the website. “I don’t think we’ve been as good as we could be about demonstrating the value-add that science and technology is bringing to the efficiency, to the transparency and to the value for money that policing’s offering the public,” Taylor says.
“I think we should only be using public money if we can demonstrate and have a track record of showing how valuable it will be,” he adds. “The challenge is to get to that point where it’s demonstrably clear that this investment will save us money in the short to medium or even longer term.”
“I think we should only be using public money if we can demonstrate and have a track record of showing how valuable it will be"
Not only is that understanding critical to knowing which areas to invest in but, Taylor hopes, it will also help to make the case for investment where resources are already stretched. He shares the example of a study that Lancashire Constabulary began in 2018 to improve the performance of its force control room by getting a clearer picture of demand and capacity.
The force knew it received 1.2 million 999 and 112 calls a year, but had very little data on what those calls were about. Using voice-to-text transcription and natural language processing enabled the force to understand why people were calling, and when the most high-risk calls were coming in. It also found a percentage of the calls were repeat calls – a problem it could tackle “right away”, Taylor says.
“Ultimately, what S&T does is save resource,” he explains. “So when people say to me, ‘I can’t invest in this, I just don’t have the resource,’ my counter argument is ‘we need to invest in this because S&T frees up that resource’. It allows officers to go back to doing the things that they really want to do.”
Taylor is preparing to kick off a programme to improve how the impact of science and technology interventions in policing is measured, which he says is currently not as developed as it should be. “We have some good news stories, and we can talk qualitatively about what’s being done, but I’d really like us to be systematic about being able to demonstrate value,” he says.
This is one of his biggest challenges – and underscores his overall goal, which is to see more effective engagement with and investment in science and technology. “In a sense, our job is ‘death by a thousand cuts’ – not death, that’s the wrong word. Achievement by a thousand cuts,” he says. “It’s the cyber expert in Hampshire who can accelerate his project so it supports local people in Southampton, which we then try and scale up nationally. It’s never just one thing. It’s about creating a system that works really, really effectively.”
Taylor on... Tackling sex-based violence
The National Police Chiefs’ Council is working with UKRI as it prepares to launch a fund for projects to tackle violence against women and girls, which will allow researchers to come into police forces as secondees to work on projects in that area.
Meanwhile, Taylor and his colleagues are working with the College of Policing to support work to tackle such violence in two main areas – filling in gaps in the behavioural science that could improve safety, and tech solutions. Behavioural science could help organisations like British Transport Police tackle assaults on transport infrastructure, he says. “Firstly, we want to make it as easy as possible for the victim to report, not just immediately, but subsequently; we also want to make it as hard as possible for a perpetrator to be able to commit an offence; and thirdly, we want to think about how a bystander might reasonably intervene.
“There is a body of literature in each of those areas, but we’re keen on trialling different ways to encourage that community ‘social good’ or bystander intervention where it is appropriate.” He likens this approach to research that has been done on suicide intervention at railway stations.
On the tech side, a number of apps are in development that scientists hope could help make public spaces safer. One example is the StreetSafe app that the Home Office launched last year, which allows members of the public to report areas where they feel unsafe. Police can then use the data to make improvements such as installing CCTV or street lighting or introducing night-time patrols.