Integrated software solution for the 3-dimensional capture and analysis of footwear evidence

Footwear impressions provide a source of evidence within a range of criminal investigations including gathering of criminal intelligence. A potential suspect will leave foot or footwear impressions en route to, at and while exiting, a crime scene. This not only allows a sequence of events to be determined but may also link a suspect to a scene if their footwear is distinctive, for example as a consequence of damage or wear. Indoors crime scene officers deal mainly with two-dimensional traces; impressions left by a foot tracking mud, blood or other bodily fluids. Three-dimensional tracks, the proverbial footprint in the flower bed, are common at outdoor scenes. Traditional methods for the collection of three-dimensional traces consist of photography and casting, supported by two-dimensional pattern recognition that can type a footwear sole to a particular make or model of shoe. Three-dimensional imaging is now available as an alternative or complementary option, particularly as algorithms for digital photogrammetry have improved dramatically in recent years allowing easy operational deployment. No expensive three-dimensional scanners are required, only that a crime scene photographer take a few moments to collect additional oblique photographs of footwear impression. Consequently, three-dimensional analyses of footwear impressions are now already possible at a routine operational level, but remain the exception rather than the norm.

As part of previous NERC grant NE/H004246/1 into ancient footprints we developed a range of methods and freeware to facilitate the three-dimensional capture and analysis of footprints. Engaging with police as part of the Impact Plan demonstrated the interest and potential of such methods to enhance the analysis of three-dimensional trace evidence especially by allowing statistical analysis of differences between tracks and/or footwear. Currently comparison is done primarily via visual inspection rather than by quantitative and statistical comparison. Working with our Project Partners we propose to employ a software engineer to draw on this research, practice and existing code to create a single integrated software application for the capture, analysis and presentation of three-dimensional footwear evidence which will allow routine operational deployment by police and forensic agencies both in the UK and overseas. This will change the fundamental cost-benefit ratio associated with the collection of this type of evidence, such that three-dimensional imaging can become the norm rather than the exception. Software of this sort needs to be available to all parties involved in forensic jurisprudence - the defence as well as the prosecution - without handicap of cost. As such our proposed software will be made available as freeware rather than commercialised, which will also assist with user adoption. This knowledge translation has the potential to contribute to criminal investigations and in the safeguarding of society.

This project arises directly from the Impact Plan associated with NERC NE/H004246/1 on ancient footprints one of the objectives of which was to raise awareness of three-dimensional imaging of tracks with law enforcement agencies. The information and feedback obtained via this route in particular from the Metropolitan Police led directly to this proposed Innovation Project. The core outcome for law enforcement and forensic agencies will be to change the cost-benefit ratio around the routine operational use of three-dimensional imaging of footwear evidence. This we believe will in turn lead to an increased use of such trace evidence in crime detection, prevention and intelligence gathering thereby benefiting society at large. This will be evidenced by rapid user adoption of the freeware in the first instance (interim impact) and secondly by feedback from the user community to document the actual impact on crime detection and prevention. A clause within the freeware license will require users to provide feedback on request for impact monitoring purposes. In terms of reach our initial target is UK police and forensic agencies, however the reach is not limited geographically and we will subsequently market and target our freeware at overseas crime and forensic agencies. Footwear evidence is currently a relatively small part of the range of trace evidence used in criminal investigations and therefore the significance will initially be modest, however by changing the fundamental cost-benefit ratio around footwear evidence our product will lead to an increased weighting given to footwear evidence in criminal investigations thereby increasing the significance and impact of our freeware over time.

We also identify the general public and school and college students as potential beneficiaries of this project, along with future generations of forensic practitioners. The product will improve education and research around footwear evidence. Additionally, we aim to use a participatory approach to sourcing the data required for the machine learning component of the project, which, although a supplement to the core product, is nonetheless an exciting opportunity to enhance the range of outcomes. The benefit and outcome here is to the wider STEM agenda and given the popular potency of forensics, participatory science and artificial intelligence, we believe that the impact here could be considerable. One of the key requirements of machine-learning algorithms is data, in this case multiple footwear tracks with placed landmarks. We will engage university students on forensic programmes at Bournemouth and Liverpool John Moores universities to collect this data. More importantly, however we will run workshops and 'field-days' at the Bournemouth University outdoor crime facility on the Trigon Estate in Dorset to engage school children in creating, capturing and analysing their own tracks thereby contributing data to the project and participating in a STEM inspiring activity. We will support this activity by a strong 'engagement' element to the projects website, supported by a social media campaign, and coordinate this activity via two Summer Undergraduate Research Assistantships (URAs) funded by Bournemouth University as a direct contribution to this project (6 weeks, Full Time). URA's are paid posts that allow students to work with academics on live research projects gaining experience for their own research, while also fuelling future research aspiration.

The outcome for those engaged in education/research into vertebrate tracks will be the provision of a community-wide platform for the collection and analysis of track data improving methodological and analytical standards and data comparability. This academic outcome will support research into ancient environments and into the biomechanics both of extant and extinct animals.