‘Science and technology is becoming more central to efficient and effective policing’

“For some time now,” says Professor Paul Taylor, “there’s been an emerging recognition that science and technology is becoming more central to efficient and effective policing. We also live in a world in which technology such as cyber and the Internet of Things, along with data and behavioural science are all part of the footplate of policing that contributes to a picture that’s very different to traditional policing of the past.”

Taylor is speaking as the UK’s first Chief Scientific Adviser (CSA) in the policing domain. The newly created post at the National Police Chiefs’ Council (NPCC) – Taylor took up the position back in May 2021 – is funded by the Home Office, a UK government ministerial department, and was, according to Taylor, “warmly welcomed by the chiefs” who had been keen on the creation of a CSA position for “some time”. The job comes with the remit to connect science and technology expertise in the UK and globally to keep policing at the fore of best practice. As CSA, Taylor will be tasked with guiding critical strategies, policies, and decisions, “helping to protect millions of people” by focusing on crime prevention.

Taylor says that to do this, a central focus of the post will be to employ emerging evidence, research and innovation in science and technology (including both data and behavioural science) to advise the police force on “opportunities and risks” associated with crime reduction. The creation of the specific CSA role is timely says Taylor, because “more than ever, today the police are clearly a huge consumer and innovator of technology, so why wouldn’t there be a CSA to try to bring some cohesion across policing?”.

According to the UK government’s website “most government departments have a Chief Scientific Adviser (CSA) to provide scientific advice”. The purpose of this network is to advise the government CSA on all aspects of policy on science and technology. In particular, they “provide advice to ministers, through the Cabinet committee system: discuss and facilitate implementation of policy on science, technology, engineering and mathematics (STEM); identify and share good practice in STEM-related areas, including the use of scientific advice in policy making; and facilitate communication on particular high-profile STEM-related issues and those posing new challenges for government”. The fact that the NPCC has been without a CSA until now – despite an internal ‘burning desire’ for one – Taylor attributes to being “an historic thing. Most CSAs are associated with government departments, and the NPCC isn’t a government department in the traditional sense.”

Taylor describes himself as a scientist that has “oscillated between working in practice and academia. I like the intersection between the two.” He says one of the reasons for preferring to have a foot in both camps is that working in the field means “you get a clearer picture of what the actual challenges that science and technology can solve are”, which in turn “focuses your academic research”. A second reason is that “when you’re designing technology that is being used in policing, national security or whatever domain, you have an outcome measure”, by which he means that “you can see when you’ve achieved something”. He says working at this intersection means he’s able to “translate one to the other. Technology quite often doesn’t make it across ‘the gap’ as it is sometimes called. That translation is something I’ve spent a lot of my career doing – getting the academic research through to the actual practitioners on the ground.”

Taylor is Professor of Psychology at Lancaster University, whose academic interest is in behavioural science. He maintains this ‘foothold’ on his academic career simultaneously with his post at NPCC and describes his three-year appointment as “a job” rather than a secondment. This type of arrangement, says Taylor, is the norm for the role of CSA, which is normally a three-year position. He says this duration limitation, combined with the fact that “you’ve got somewhere to scurry back to at the end of it”, means CSAs are more likely to feel they can make statements that run the risk of being unpopular, controversial or simply not consistent with the government’s public position. This brings to mind some high-profile televised conflicts that advisers such as Patrick Vallance and Chris Whitty have experienced during the coronavirus pandemic of 2020-21. Taylor concedes that until the recent public health crisis, “people probably didn’t realise there even were science advisors”.

Under Taylor’s watch, a new science and technology strategy for policing will be developed alongside the delivery of the National Crime and Justice Laboratory that has the overall aim of ‘transforming’ the prevention and reduction of crime. He is also there to bring together disparate functions and aspects of policing: “I’ve been continuously impressed by how much innovation is going on in different forces in different areas. Yet in terms of who knows what, it is very much a fractured landscape.” What this means, says Taylor, is that one of the first things he needs to do as CSA is “to get a full understanding of what’s going on and what our capabilities are. Finding out what the risks and opportunities are. Making sure we continue to develop and go forward by engaging academia and industry. So there needs to be a much more crystallised picture of the current landscape.”

He is also, despite having the responsibility for only three years, looking at where policing needs to be in “15-20 years, working out the role technology will play in that landscape. The chief constables have signed off several thematic top-level areas where they suspect there is great value for policing. So, my first job is to draw the map of how we get there. More importantly, it’s to work out how that is going to be achieved. What’s the practical nature of making it happen, because we’re not going to be able to do it alone. Policing needs to engage all science and technology of the UK system. For example, low TRLs [technology readiness levels] working with academia. Then, when we have more mature products, working with industry and other professionals to pull those into workable outcomes. But we also need to work across the CSA’s network. Already, within my first few months, I’ve begun to see opportunities for cross-learning with other departments.”

The ultimate objective of the NPCC is “to assist with creating a safer society”, says Taylor who is keen to break down what this means for members of the urban public in particular, cynical over how the fine words of yet another top-level official will get drug dealing, theft, violent crime, prostitution and antisocial behaviour out of their neighbourhoods. “If you take this scenario as an example, then as CSA you’ve got to look at how to develop routing algorithms and other sophisticated ways to make sure that our police are at the hotspots that they need to be at and at the right time. Then you need to look at whether you can use other organisations to support that. If it’s happening around, say, a tube station, you need to see if you can work with Transport for London. We’re doing a lot of work with the British Transport Police around indecent behaviour on transport and how to tackle that from a scientific perspective, which brings in behavioural science. Our job is prevention. We don’t really want to be in a reactive space making arrests.”

At this point, Taylor cites the example of the Thames Valley Violence Reduction Unit that has “working relationships with lots of partners in the region, from schools to local authorities, and they have data from each of those organisations that gets put into a data lake, that then provides them with a picture of people who are potentially vulnerable, which they can then act on.” We then take as an example the profile of a hypothetical young drug dealer, truant from school for days, father in prison, unemployed mother “and so on. We identify this as a risk case: we can then ask what we can do to put safeguarding in place for that person. It might not be police safeguarding. It might be that you work with the school or community support. So, there is a wider sphere of working with others.”

This type of intervention is driven by data, says Taylor, and he puts it to the first national policing CSA that the public is heartily disenchanted by the ‘driven by data’ mantra with which it has become so familiar during the government’s coronavirus pandemic televised press conferences. Taylor accepts that the term ‘data’ doesn’t necessarily have a positive resonance with a public equally disenchanted by what they perceive as being a pandemic of street-level crime. “But the point of using data is to create efficiencies. It won’t eliminate the problems, but it will create a drop in them.” Recalling a “great cradle to grave” example of how data can create efficiencies, Taylor explains in broad figures about a force that is struggling with a volume of 1-1.2 million 999 calls per month. “Now, historically, all we’ve ever been able to say about a situation like that is maybe to break it down into the number of calls per day. It’s a lot of calls and it’s problematic because until now we’ve not really been able to do anything clever with that to reduce the problem.”

Efficiencies have been created in such scenarios with the introduction of voice-to-text technology to transcribe calls, which are then analysed using natural language processing to pull out key words so that calls can be assigned degrees of urgency, or ‘triaged’. “There are sets of words that tend to indicate mental health concerns, for example, and the calls can get put through to the right person. This is where what is well-established tech is being used in an innovative space to create some incredible efficiencies. This was all done manually before, and now at least in the first triage it is being done automatically, allowing the police to refocus their activities. There are huge gains to be made in terms of time being freed up and improved service,” in the sense that the public-facing response is “simply better”.

Data use in supporting human decision comes with balancing pressures “that policing needs to get right. We do not want to get into the situation where a police officer is trusting a machine-learning algorithm’s prediction beyond his or her own intuition.” But on the other hand, neither do “we want the algorithm to be batted away as irrelevant because it doesn’t fit with that intuition. That integration is a complex problem that needs to be solved.” This fits in with an area of Taylor’s academic expertise in which he is involved with “quantifying the qualitative”. “There are lots of behavioural challenges out there. If we could quantify that behaviour – digital footprints are a good example of that – then we can start to use some of what data science and technology has provided to help us analyse some of that softer behaviour. This could be used in triaging risk assessment in counterterrorism. Some judgement tools we use focus on static and dynamic risk factors, such as demographic and life-stage details. We can quantify those and use them in a structured professional judgement about where to focus our resources. The technologies are there to assist a decision maker.”

Taylor says perhaps the biggest eye-opener on taking up his post as the NPCC’s CSA is just how widespread the potential for automation is in policing. “There are currently lots of jobs still being done manually which could be automated. This could just be as simple as form filling. In fact, some of the NPCC’s work with the Accelerated Capability Environment has been around getting teams in and helping us with that automation.” It’s a quick win, reckons Taylor, “because it relieves officers of so much time-consuming paperwork”. Yet the real game-changer will be the intersection of data and digital technology.

“The National Crime and Justice Laboratory is an example of us pulling together the various data sets that sit across the criminal justice system. Not just in policing, but in justice and prisons, so that we can better understand an offender’s journey through the process. Beyond that, pulling together data from schools, local authorities and so on to create a more holistic picture will allow us to see where the system has got snags, so that ministers can make much more informed decisions about where funding should go. But it will also help forces, because if we have a national picture of what is going on, we will begin to start routinely to do really effective trials of different technologies.”