Guidance

Market Exploration Document: Facial Recognition

Updated 5 October 2023

The Defence and Security Accelerator (DASA) is running a Market Exploration to find technological solutions for the use of Facial Recognition within policing and other security agencies.

1. Summary

On behalf of the Home Office, DASA is running a Market Exploration to identify higher Technology Readiness Level (TRL) or adjacent capabilities that could be implemented within the next 18 months. The aim of this exploration is to help the Home Office to have a better understanding of current technologies that could enhance Facial Recognition capabilities.

Your submission will be provided to the Home Office and be visible to law enforcement and security stakeholders within government. Technologies and concepts may be adopted within any future acquisition programmes (the 43 Home Office police forces of England and Wales are an example of potential customers for such technologies).

Please note that this request for information is not a commitment to subsequently launch a formal DASA competition.

2. Background

Facial Recognition (FR) technology can be used in a number of ways within policing and security settings to prevent and detect crime, protect borders, find wanted criminals, safeguard vulnerable people, and to protect people from harm.

As an increasingly important capability for law enforcement and the Home Office, there are significant opportunities to improve public safety through the use of FR. Explaining how FR works and how it will be deployed in a way which maintains public trust is critical and complex. In increasing the use of such technologies, the Home Office is committed to:

• Explaining how FR works, and how it protects the public in combination with other tools and technologies.
• Explaining who decides where and how to use FR and how they know it is safe, effective, lawful, ethical and value for money.
• Ensuring that our operational partners have the best possible data, technology and training available.

In law enforcement and security, current applications or use cases generally fall within one of the three following categories:

1. Retrospective FR (RFR), a system for identification (or one:to:many) to be used after an event to help establish who a person is or whether their image matches against other media held on image databases.

2. Operator-initiated FR (OIFR), a system for identification (or one:to:many) where an operator can decide that they need to take an image of a person and then use FR software to help them establish who that person is. This use of FR can help provide an identification of someone who is providing false details or is unable to communicate who they are.

3. Live FR (LFR), a real-time system where stand-off cameras are focused on a specific area; when people pass through that area their images are streamed directly to the Live FR system. This system contains a watchlist of wanted persons, or those who pose a risk of harm to themselves or others. The system provides intelligence alerts to operators who will make a decision whether to take further action.

The use of live systems in policing is controlled through the College of Policing Authorised Professional Practice (APP) which was published following legal scrutiny of their use. Following this practice ensures that the use of live systems is fair, proportionate and free from bias.

The use of FR technologies is varied across UK policing. Improving the effectiveness and consistency of its use is a priority for the Minister for Crime, Policing and Fire and the Policing Chief Scientific Advisor (CSA).

3. What we want

We are interested in innovative capabilities

• to resolve identity focused on facial features and facial landmarks
• that support integration, algorithm development and analytics
• that support the three categories described above
• that are potentially deployable over the next 12-18 months.

We are seeking innovations focused on multiple areas of FR including biometrics systems (e.g., algorithms), wider systems integration (e.g., integrating the algorithms with front-line technologies) as well as associated technology areas such as CCTV and video analytics.
Submissions should be provided by teams with the experience and knowledge necessary to establish sound scientific evidence for any proposed technology. It is vital that such technologies are secure, accurate, explainable and free from bias.

By completing the Capability Submission Form neither DASA, the Home Office nor yourselves are committing to anything, but your submissions will be compiled by DASA and provided to policing and Home Office partners for their due consideration. Your submission will also help us to identify your interests in this area.

4. What we don’t want

We are not interested in technologies beyond the resolution of identity through FR, such as

• Iris detection, gait analysis and/or other biometric modalities.
• Hardware (without associated algorithms and analytics).
• Object and/or anomaly detection.
• Analysis/ determination of physical characteristics (e.g. age detection).
• Analysis of behavioural characteristics (e.g. lie detection).
• Literature reviews, paper-based studies, consultancy.
• Non-technical solutions or marginal improvements to existing capabilities.

This is not a competition and therefore we are not asking for costed proposals at this stage. This is a market engagement request for information exercise and we do not commit to subsequently launch a DASA competition.

5. How to submit a Market Exploration Submission to DASA

Responses to this Market Exploration must be submitted via the DASA submission service, for which you will need to register. We recommend you use a Google Chrome browser to access the DASA submission service.

You will be asked for a title and short summary of your innovation, along with questions related to your organisation, your idea and technology maturity. We are seeking to understand what and how much further development is required for a complete solution to meet requirements, or whether a combination of separate solutions is required. The information you provide may assist in developing a statement of requirements for potential future activities.

Early submissions to this exploration would be welcomed. DASA will share received responses with the Home Office on 12th September and 26th September. The final deadline for submissions will be midday (BST) on 17th October. Unfortunately, we are unable to accept any submissions after this point.

Please only provide details of one product/capability per submission. If you have a number of potential solutions, then please submit multiple forms. Please include the type of FR technology that you are attempting to address in your title/s.

If you have any questions then please email accelerator@dstl.gov.uk with” Facial Recognition market exploration” in the subject line.

6. How we use your information

Information you provide to us in a Market Exploration Submission, that is not already available to us from other sources, will be handled in-confidence. By submitting a Market Exploration Submission you are giving us permission to keep and use the information for our internal purposes, and to provide the information onwards, in-confidence, within UK Government. The Defence and Security Accelerator will not use or disclose the information for any other purpose, without first requesting permission to do so.