When All is Tread and Done
The evidential use of footwear in criminal investigations has long centred on outsole impressions; tread marks recovered from crime scenes and compared against seized shoes. Yet in recent years, a new dimension of footwear evidence has emerged: the upper portion of shoes (known as uppers), as captured in CCTV, body-worn cameras, and mobile footage. These images, previously dismissed as incidental, are now being recognised for their forensic potential.
With a view to driving world-leading innovation, but also efficiency and cost-savings for local forces, a team of subject matter experts from West Yorkshire police, specifically Yorkshire and the Humber Regional Scientific Support Services [YatH RSSS] alongside academics from Staffordshire University have delivered a remarkably agile three-phased research and development project. National rollout is now on the horizon.
The project aimed to transform footwear uppers from overlooked visual cues into a validated and operational form of forensic evidence.
When asked about the depth and speed of the project Dr. Claire Gwinnett, Professor in Forensic and Environmental Science at Staffordshire University, said: “Over the course of these three phases we’ve gone from idea, to research, to operationalising. We had a perfect point in time for this, there’s more phones, there’s more CCTV. We had the expertise and the support. Innovation can be fast, it doesn’t need to be researched for 20 years.”
Claire is also the director for the Centre for Crime, Justice and Security at Staffordshire University, across multiple research centres. In discussing the research. She’s joined by Ryan Harris, the footwear technical manager for Yorkshire and Humber, who has decades of experience in forensic science, including with the Home Office Forensic Science Service with a specialism in digital imaging. Along with Selina Reidy who joined West Yorkshire Police in 2006, where she specialised in undertaking footwear mark coding and comparisons for both volume and major crime cases and later qualified as a podiatrist - joining the Forensic Unit at Sheffield Teaching Hospitals NHS Foundation Trust until 2020.
In delivering their research, the team moved at speed, while ensuring quality standards, ethical considerations and stakeholder engagement. The Office of the Police Chief Scientific Adviser funded the multi-phase research and implementation initiative titled “When All Is Tread and Done” via the Police STAR Fund. The fund is an annual innovation call which aims to stimulate local innovation and encourage collaboration to solve science and technology problems within policing.
This project has demonstrated the growing forensic value of footwear upper comparisons, responding to increased demand with a structured, evidence-based approach.
The three phases of the “When All Is Tread and Done” research programme tackled different aspects: developing a reliable method for capturing and cataloguing footwear uppers under forensic conditions, establishing the evidential value of footwear upper features through controlled research, and finally, operationalising the capability nationally through tools, training, and policy integration.
Image source: West Yorkshire Police
Strategic Context and Operational Need
Why footwear uppers? Between 2018 and 2023, the Yorkshire and Humber Regional Scientific Support Services observed a marked increase in casework involving footwear uppers. By 2022–23, evidential comparisons had risen by 25%, despite the absence of formal promotion or infrastructure. Other forces reported similar growth.
Investigators were increasingly relying on visible footwear in footage to identify suspects, link scenes, and corroborate witness accounts. However, the process lacked standardisation.
There was no national database of footwear uppers, no validated imaging method, and no research base to support expert conclusions. The evidential strength of such comparisons remained uncertain, and their admissibility in court was limited.
The idea to begin a shoe upper database as a new form of evidence was officer led, says Selina Reidy, noting that, ‘CCTV was being collected when officers attended scenes, even if CSI weren’t in attendance. Investigating officers might notice footwear on the footage and get in touch with us to see if we could help. Requests came in regionally, but we also had requests from the NCA.” The demand for a technical solution was clear.
Scientific Foundations and Key Findings
The project’s first phase focused on technical feasibility. Researchers developed a cost-effective imaging method using a consumer-grade CCTV camera and motorised turntable, capturing 360° views of footwear under both white light and near-infrared (NIR) illumination. This setup replicated the conditions under which shoes are typically recorded in surveillance footage.
A pilot database was created using images of popular shoe models. Early findings revealed significant variability in how shoes appeared under NIR. In one instance, the left and right shoes of the same pair exhibited different stitching visibility, raising questions about material behaviour and evidential reliability. These anomalies underscored the need for deeper investigation.
Phase two set out to answer a critical question: can the visible features of footwear uppers stand up as forensic evidence?
To find out, researchers ran a series of controlled studies. In the visibility trials, shoes marked with blood stains, scuffs, and creases were tested under varying lighting and environmental conditions. Blood and holes proved remarkably resilient, detectable at distances of up to six metres, while creases were more elusive. Factors like shoe colour, reflectivity, and light source played a significant role in what could be seen and from how far. The team experimented and built their own tools, or re-purposed existing products. They found infrared inspection binoculars and made minor adjustments – using in-house 3D printing – and developed the product into a perfectly tailored tool. This approach also saved money.
Next came the wear trial. Over nine months, volunteers wore shoes as normal while researchers tracked the evolution of features like creases, dents, frays, and stains. These marks persisted, deepened, and in some cases transformed, as creases became structural damage and frays turned into holes. Dirt, by contrast, faded quickly. The study confirmed that many upper features are not only unique but stable over time.
To understand how common these features are, the team imaged 1,000 shoes and catalogued over 7,600 individual characteristics. An interactive dashboard was built to help practitioners interpret these patterns and assess the significance of what they see in footage.Finally, material and dye analysis revealed that infrared visibility isn’t just about fabric, it’s about chemistry. Some black shoes appeared white under NIR (near infrared) light due to disperse dyes, while pigment-based colouration held its tone. These findings have direct implications for interpreting surveillance footage and predicting shoe colourways.
Together, these studies laid the scientific foundation for footwear upper comparisons, giving practitioners the tools to assess visibility, persistence, and rarity with confidence.
The project produced several impactful outputs, including a novel image quality assessment tool, an interactive footwear characteristics database, and insights into the persistence and commonality of wear features.
Image source: West Yorkshire Police
National Integration: From Research to Routine Practice
With the evidential value of footwear uppers firmly established, the final phase of the project shifted focus from research to real-world impact, ensuring this capability becomes a routine part of forensic practice across UK policing.
The priority is integration. That means embedding footwear upper comparisons into everyday workflows, equipping practitioners with the tools to extract intelligence from surveillance footage, and providing the training needed to interpret and report findings with confidence.
One of the most transformative developments the team is delivering is a machine learning tool designed to identify shoe make and model from CCTV images. Built on thousands of reference images and enhanced with advanced re-colouration and image correction techniques, the tool is built to perform even under poor footage conditions, supporting fast, reliable intelligence generation. Given CCTV is often blurry and pixellated this is a key requirement.
Ryan reflects from a policing perspective: “Machine learning does two things, one of them is efficiency, because at the moment we would be restricted to manual searches. We’re hoping that the machine learning, once we drop the image in quickly brings back those results to improve efficiency. And the other thing to touch on is the accuracy - we’re not just relying on the machine learning to give us these answers. The idea is it will be similar to fingerprints so that the machine learning brings back a lot of options that a human will then go through and look at similarities and differences – ensuring speed, accuracy and quality.”
Training will be rolled out nationally, with examiners receiving both online and in-person instruction, backed by competency assessments aligned with the Forensic Science Regulator’s Codes of Practice.
To sustain momentum, the project has launched the Shoe Upper Network (SUN), a national community of practice that supports ongoing research, casework collaboration, and database expansion. SUN ensures the capability continues to evolve in response to operational needs.
The evidential framework is also being strengthened. Work is underway to develop guidance on combining upper and tread evidence in reports, and to explore how cognitive bias might affect image-based assessments. These efforts aim to ensure that footwear upper evidence is not only scientifically robust, but also courtroom-ready.
The initiative has earned strong support across NPCC portfolios, including Forensics, Digital Policing, and Artificial Intelligence, and has been endorsed by the National Fingerprint and Footwear Board and the FINDS Strategy Board. With peer-reviewed publications, dissemination events, and international collaborations planned throughout 2025–26, the project is now moving decisively from research into routine practice.
“When All Is Tread and Done” has transformed an underutilised visual cue into a validated forensic capability, supported by rigorous science and national coordination.
Dr. Claire Gwinnett, Professor in Forensic and Environmental Science at Staffordshire University, noted: “It is quite rare to have an idea and no evidence base, and to move so quickly to it being put into operation. Our work across the three phases has gone from idea, to research, to operationalising. It’s proof that that we can be fast and accurate. Together we are creating ground truth data.”
Selina Reidy and Ryan Harris from West Yorkshire police agree. Selina adds: “We have the data. We were data rich but time poor, so we needed the academic input – it’s very helpful to have that partnership. This will be valuable to forensics and investigations nationally. There are also international opportunities to share this approach.”
Ryan recognises the wider support: “We never under-estimate the support for innovation we’ve had from our SLT (senior leadership team) here in West Yorkshire police.”
Footwear uppers, once incidental in surveillance footage, are now a source of identification, intelligence, and evidential strength. With tools, training, and standards in place, this capability is being operationalised across UK policing, enhancing investigations, strengthening court outcomes, and advancing the science of forensic identification.
As the project enters its final phase, its impact is already being felt. Forces are adopting the imaging method, contributing to the database, and preparing to integrate footwear upper comparisons into routine casework. The NPCC’s strategic Police STAR investment has delivered a new evidential standard - one that will shape the future of forensic practice for years to come.
