Brian Mitchell

Dear Sir Inquest touching the death of Brian Mitchell I write on behalf of Transport for London (TfL) and the Mayor of London regarding the Senior Coroner’s Regulation 28 Prevention of Future Deaths (PFD) report dated 29 December 2025 following the inquest touching the death of Brian Mitchell which took place on 15-17 December 2025. I would like to take this opportunity to repeat, on behalf of everyone at TfL, our sincere condolences to the family and friends of Mr Mitchell for their tragic loss. Prevention of Future Deaths (PFD) report After the inquest, the Senior Coroner sent a PFD report to TfL, the Mayor of London and the Department for Transport (DfT) raising four matters of concern. By way of background, the Greater London Authority (GLA) consists of the Mayor of London and the London Assembly. The executive and decision-making responsibilities of the GLA are vested in the Mayor, whilst the London Assembly scrutinises the work of the Mayor through its committees and plenary sessions. TfL is a functional body of the GLA and runs most of London’s public transport services. The Mayor of London is the Chair of the TfL Board and appoints the Board which appoints the Commissioner of TfL. TfL exercises its functions, in part, for the purpose of securing or facilitating the implementation of the Mayor’s Transport Strategy. TfL has three subsidiary companies, one of which is Transport Trading Limited (TTL). TTL is the holding company for all of TfL’s operational delivery companies. London Underground Limited (LUL) is a wholly owned subsidiary of TTL. LUL is responsible for operating the London Underground (LU) train network. The Office for Rail and Road (ORR) is the regulator for LUL’s services and is also responsible for monitoring implementation by LUL of any recommendations from the Rail Accident Branch (RAIB). 1

This response is sent on behalf of TfL and the Mayor of London both of whom are grateful to the Senior Coroner for raising these concerns. We set out below a response to each of these concerns. Tb® cireumstemces of Mir Mifchell’s death The PFD report sets out the Senior Coroner’s summary of the circumstances of Mr Mitchell’s death in section 4. The inquest into Mr Mitchell’s death followed a detailed investigation by the RAIB who published a report in January 2025: The inquest was conducted in accordance with the principles established in the case of R (Secretary of State for Transport) v HM Senior Coroner for Norfolk [2016] EWHC 2279 Admin. This meant that the Court was bound to the findings of the RAIB as to the causes of this tragic incident. We note that the RAIB found that Automatic Train Operation (ATO) was a possible underlying factor, in Mr Mitchell’s death, rather than a likely cause. We also note that the RAIB did not find that the initial collision with Mr Mitchell was likely to have been avoidable. The RAIB did find that LU had not fully quantified the risk of a passenger falling from the platform and being struck by a train at Stratford station, and its local risk controls were not sufficiently effective to prevent this incident. Response to the Coroner’s areas off concern
1. Sn the two years that have elapsed sone® Brian’s death investigations have been undertaken by the British Transport Police, the Rail Accident investigation Branch (RA1B) and TfL into the circumstances that Bed to this incident. There is no ciear evidence to demonstrate that risks of fatal! harm have been mitigated. We recognise that there were opportunities which had the potential to prevent this tragic incident, including: ° the bag which was on the platform was an opportunity to have potentially identified that Mr Mitchell had fallen on the track; ° recognising that Mr Mitchell had been sitting on the platform for an extended period of time; and ° ensuring our train operators have the right skills and tools to maintain concentration as they approach a platform. We have taken clear action so that, should a similar incident occur, our station staff and train operators are more likely to identify unattended property and potentially the length of time a customer may be waiting on a platform. This includes changes to our operational procedures and staff training, which are detailed below. Our approach to improving safety has included increasing when station colleagues carry out visual checks on stations to prevent infringement of the track areas and Platform Train Interface (PTI) incidents. A video has been produced which contains a briefing for station colleagues reminding them of the importance of completing thorough station security checks including an emphasis on customer service and being a visible presence on the station. This was launched in June 2024, and 3719 (85%) station colleagues have now been briefed including all station colleagues at Stratford station. The changes relating to these checks have been formalised in TfL’s operational safety rules and 2

training, ensuring the learning from this incident are shared with station colleagues now and in the future. We have also updated our security rules for station colleagues and train operators to help them identify whether lost property is suspicious or not. Our rules now require our front-line teams to visually check the track for unauthorised persons (once an item has been declared non-suspicious) when items are discovered near the track (including platforms and overbridges). This also includes appropriate checking when a train is berthed in a platform. All stations, trains and service control colleagues had been fully briefed on these rules by 12 May 2025 and the Rule Book and associated training updated in line with the change. The change has also been incorporated into training for new station colleagues and train operators. While recognising the benefits of providing seating for customers in our stations, we are finalising plans to remove the specific bench on which Mr Mitchell sat from this location, given that it is a location which is not regularly frequented by customers and there are other benches on this platform, which are less remote from passage of other customers and staff. We have started plans to remove the bench and this will be completed by 6th March. These steps will hopefully mitigate the risk of a similar incident happening at Stratford station. We recognised at the time that the mitigations needed to further reduce this risk are not readily available to the rail industry. We consider that technology has a role to play in, for example, detecting someone on the track which would allow us to take action to stop train movements. We have set out details below on the action we have taken, and our plans, in our response to the Coroner’s second area of concern. TfL’s Vulnerable Customer Programme Safeguarding is a core organisational priority for TfL. Our TfL Strategy sets out our commitment to enhance our safeguarding response to protect our most vulnerable customers, identifying and mitigating risks. This includes managing the behaviour and safe travel of young people; reducing the risk of suicide and supporting customers experiencing mental health crises; addressing rough sleeping; preventing exploitation; and responding to emerging forms of temporary vulnerability, including intoxication. In 2023, our approach to safeguarding on the LU was recognised through accreditation under the DfT’s Safeguarding on Rail scheme, with similar practices embedded across our other transport modes. Although we do not hold a statutory safeguarding duty in most circumstances, we recognise the significant role we, as an integrated transport authority, can play in preventing vulnerable children and adults from coming to harm when using our services. We are committed to using our position, reach, and our daily interactions with customers to contribute meaningfully to the overall safety and wellbeing of our customers. I have set out details on the action we have taken, and our plans, below. Our internal investigations into some of the fatal or serious injury incidents on the LU network in recent years highlighted intoxication, poor mental health and potential suicidal intent as recurring factors with clear recommendations to improve our organisational response. In many of these incidents the temporary vulnerability of our customers - through intoxication or mental health crisis - was a common factor. However, data analysis of incidents has shown there is no consistent pattern in terms of the locations or 3

times these incidents will happen. This highlights the difficulty of targeting interventions to specific sites or circumstances. As a result, we recognised the need for a systemic, organisation-wide approach to safeguarding vulnerable customers, with a greater focus on intoxication and mental health crisis, to ensure that those who are temporarily vulnerable can travel safely. We have made clear our commitment to reducing the risk of vulnerable customers coming to harm through improvement in staff awareness, competency training and updating operating procedures. The TfL Vulnerable Customers Programme, which was established in April 2025, has five workstreams: ■ Workstream 1: Working with all TfL types of transport to review existing risks to vulnerable customers and reviewing wider research ° Workstream 2: Clearly defining vulnerability for our frontline teams and how we want them to respond to vulnerable customers ■ Workstream 3: Creating new training content using the outcomes of Workstream 2, updating our processes and procedures ° Workstream 4: Updating our Passenger Help Point (PHP) signage and rerouting calls to one of LU’s control centres a Workstream S: Trials on LU stations with Customer Operations and Security, Policing & Enforcement (SPE) teams working together and assessing the impact of Workstreams 1-4, and from that making recommendations about new ways of working to keep vulnerable customers safe I have set out details on action taken to date and next steps on each workstream below. ° Workstream 1: Working with all TfL modes to review existing risks to vulnerable customers and reviewing wider research In August 2025, we completed a comprehensive assessment of current safeguarding risks and mitigations across all TfL’s public transport modes. This included analysis of data and insight from research and benchmarking to help prioritise interventions by time/place/setting. Based on that work, and discussions with London’s emergency services to understand their approach, we have adapted the ABCDE vulnerability assessment framework (the framework used by the emergency services in the UK) for use on the LU network. In October 2025 we started trialling this framework across 21 LU stations (a mix of inner-city, suburban, gateway and outlying locations) to understand the effectiveness of this approach, and we plan to integrate what we learn from this trial into LU’s safety procedures by 31 March 2026. Feedback on the framework has been positive and it builds effectively on our existing suicide prevention procedures. We will commence briefings and training for operational colleagues on the framework from April 2026. ° Workstream 2: Clearly defining vulnerability for our frontline teams amd how we want them to respond to vulnerable customers in September 2025, we developed and implemented a communications and culture plan for customer safeguarding on the LU network. This has included: o Customer service principles, in particular, use of the already existing ‘protector’ principle to clearly explain and promote the role that operational customer-facing staff have in identifying and responding to vulnerability risks including intoxication and mental health issues. 4

o Developing a plan for colleague communications to raise awareness of customer safeguarding and the role of operational customer-facing staff in supporting vulnerable customers and raising awareness of the support available for staff who have been involved in managing stressful and traumatic situations. o Expanding our TfL Safeguarding Award recognition scheme and police commendations to recognise staff actions in supporting customers who are in mental health crisis or at risk of harm due to intoxication or other vulnerability. As outlined in workstream 1 above, we have adapted the ABCDE vulnerability assessment framework used by the emergency services. Our staff training content and briefing/communications materials were updated to reflect this at the end of December 2025. Our next steps are to brief operational colleagues (approximately 11,000 individuals) on this between 31 January to 1 October 2026 and update our operational rules by October 2026. Briefing our operational colleagues on the changes needs to take place before the rules are updated to ensure that they are aware of the actions to be taken. Workstream 3: Creating new training content using the outcomes of Workstream 2, updating our processes and procedures We are in the process of developing new safeguarding and vulnerability training which will be completed by the end of March 2026. This will then become required learning for key operational colleagues. This will go live by the end of April 2026. It takes a year to train our operational colleagues in line with their training cycle so this will be complete by April 2027. Workstream 4: Updating our PHP signage and rerouting calls to one of LU’s Control Centres We put new signage in place on all PHPs at all LU stations (completed 31 December 2025). Now, apart from a very small number of help points on our network where we were not able to change the call routing system due to the age and complexity of the communications equipment associated with these PHPs, all emergency calls by customers will be rerouted from all identified stations to one of our control centres if station colleagues are unable to respond to them in time. This will allow our control centres to respond quickly in the event of a customer using the PHP to raise a safety concern. Workstream 5: Trials ©Di LU stations with Customer Operations and SPE teams working together and assessing the impact of Workstreams 1-4, and from that make recommendations about new ways of working to keep vulnerable customers safe In October 2025 we started a trial on 20 LU stations on the east end of the Central line to assess the feasibility and potential impact of additional interventions to reduce the risk of customers coming to harm when intoxicated or vulnerable for other reasons, building on the foundational response of awareness raising and training. This includes: o Increased physical station staff checks of these stations between 22:00 and 02:00 o Enhanced CCTV monitoring between 22:00 and 02:00 o Increased patrols by Travel Support Enforcement Officers on Night Tube services 5

This trial began in October 2025 and will run until the end of February 2026 with an evaluation of impact and next steps to be completed by March 2026. Any learnings and best practice will be integrated into our training. A full assessment of this trial will be produced in March 2026 once it is completed. An initial review for November 2025 (the first full trial month) suggests that the scheduled checks by station colleagues provided a marginal increase in detecting vulnerable people. Most vulnerable people within the 22:00 and 02:00 time frame were successfully identified by station colleagues outside of the scheduled checks. This suggests that briefing station colleagues on the new framework to encourage best practice and building a culture of vigilance and proactive response generally through training would be effective. Travel Support Enforcement Officers reported supporting numerous customers during on-train patrols during the trial, and the benefits of this approach will be reviewed further. A further trial assessing the benefit of 2-hourly scheduled checks throughout the traffic day is planned to begin in March 2026. As well as the measures identified above, since the inquest, we have reflected on the controls in place at Stratford station. We have reviewed incidents of falls on track at terminus platforms (such as platform 13 at Stratford station). This has not identified any location specific trends in relation to incidents of this nature. This is reflective of the analysis we have also undertaken at all platforms (e.g. terminus and non-terminus platforms). Incidents by platform are low level and can occur anywhere on the LU network without platform edge doors. As London’s integrated transport body, our responsibility is to manage risk to as low as reasonably practical. Our ambition is to eliminate death and serious injury from London’s transport network by 2041. We take this responsibility seriously and actively work to go beyond that to achieve our Vision Zero goal through exploring new approaches to manage risk as outlined above.
2. Recommended technofcgicaJ measures to detect and atert staff to the presence of persons on the tracks have not been omptemented at Stratford station We shared information on TfL’s PTI New Technology Programme in TfL’s witness evidence for the inquest. We are actively trialling technology to allow us to detect someone in the track environment - whether they fall unintentionally or access the track deliberately. I have set out an update on the PTi new technology trials below. As noted in TfL’s witness evidence for the inquest, in early 2025, we created a programme to trial technologies to improve customer safety at the PTI. The trials are focused on detecting customers who: o Access the tracks - deliberately or accidentally o Fall between the train and the platform o Are caught in train doors Our three trials will test a mix of CCTV and sensor-based solutions on three different services (Central line, Piccadilly line and DLR). This technology aims to detect the specific scenarios noted above and will then send an alert to our operational colleagues to respond and reduce the potential for someone getting seriously injured. This technology also has the potential to reduce the risk of unauthorised track access incidents. 6

In December 2025, we started our first trial of the new technology - a trial on the Central line aimed at detecting someone falling or accessing the track. The trial, which is taking place on three stations on the Central Line (West Ruislip, East Acton and Mile End stations), uses a combination of new video analytics and existing camera technology to detect people who access the track environment either intentionally or unintentionally. The trial will happen in two phases, a shadow trial, followed by an operational trial. 0 The shadow trial, which commenced in December 2025, is testing the effectiveness, accuracy, and reliability of the new technology. During this time, there will be no operational responses to incidents identified by the system as the technology will be picking up false alerts as it learns what is correct.
• The length of the shadow trial depends on how long it takes to refine the technology to accurately identify PTI incidents. Once we’ve established confidence in the accuracy and reliability of the technology, we will progress to an operational trial which will allow us to respond to any incidents. If the trials are successful, we plan to use the technology more widely across our network to further enhance the safety of our customers travelling on our network. To inform the development of our trials, we conducted research into PTI technology solutions across 10 metros, including Singapore, Sydney and Delhi. We met remotely with these transport operators to understand the range of solutions that have been tested both across the UK and internationally, and to capture lessons learnt. As our own trials progress, we will continue to engage with our national and international counterparts to share insights and ensure we incorporate emerging best practice and learning into our approach. A small number of other railway operators around the world are using technology to identify people on the track. Most of these operators are managing smaller metro systems, in environments with fewer customers and using new digital cameras. Some have deployed it in a site-specific way, rather than across the network. TfL has done similarly - in response to a serious incident on the DLR, we installed technology at DLR Custom House station which is identifying people in the track area and allowing the control team to respond quickly to protect that person as far as possible. Our LU trial aims to apply this new technology to a busier metro network and using older cameras. There are thousands of cameras on the LU network (our bigger stations will have a few hundred cameras each), so we considered how we could most quickly start to use technology to help us manage risk. Stations such as Stratford have analogue (older) cameras, rather than newer digital cameras. Rather than replacing thousands of analogue cameras across 272 LU stations with digital cameras, we are trialling video analytics on our existing, older cameras. This would allow us to roll out safety improvements far more quickly. As far as we are aware, no other metros are actively using the older analogue cameras to improve customer safety. Our trial programme is also considering the use of new digital CCTV cameras. Carrying out trials using both digital and analogue cameras will allow us to identify the most effective approach for TfL in the future. In deciding on the location for the trial, we considered factors such as the locations where customers fall on track. Our data and analysis showed that approximately half of the falls on track on the LU network can happen on any station on the network. In the 7

past 7 years, there have been 2 incidents (including the tragic incident involving Mr Mitchell) where a customer fell from platform 13 at Stratford station onto the track. There are several other stations across the LU network with a similar risk profile. As a result, the locations for the trial are stations where the camera and communications technology available at the station were appropriate to the trial, and reflected different environments (e.g. above ground, below ground) that we could physically test. The technology being trialled to detect incidents where a customer has fallen on the track is innovative in the national and global railway industry. It will take time to develop the technology so that it is accurate and reliable. As part of TfL’s commitment to safety, we have established, and are investing significantly in, a focused programme to identify the technology which will help us detect these incidents quickly. Our approach is to trial several different options - sensors, station CCTV, train operator cameras and video analytics to identify an option that we can roll out as quickly as possible across the TfL rail network.
3. N© ©Hear data is available to demonstrate that training provided to train operators (drivers) to ensure that they concentrate and look at the tracks before them whilst operating trains using ATO has resulted in positive improvement in performance Following this tragic incident at Stratford station, we recognised the need to improve training for train operators. During 2024, we updated our training and started delivering this training in November 2024. It takes a year to completely refresh and train our existing 3500 train operators, and delivery of this training was completed in December
2025. New train operators are also receiving this as part of their training. Accidental falls on track are not frequent on the LU network, and it is not possible to quantify the impact of the training in absolute terms. However, we have considered the following in gauging the effectiveness of our training on our 3500 train operators. Our Operational Training and Skills Development Team have completed a review of in- service competence assessments of train operators over the last 12 months. This has provided us with evidence that train operators are observing both the PTI and the track when they are berthing in platforms. As a result of this review, we have identified further improvements relating to how these assessments are recorded and by the end of April 2026, we will provide further advice and guidance to our internal competence assessors. This will strengthen the focus on PTI and track observations during train operator competence assessments. Monitoring of assessments will continue as part of our business-as-usual processes and a more detailed audit on the effectiveness of the approach we are taking on in-service competence assessments will be completed by September 2026. Following this, we will review the effectiveness and make further changes as needed.
4. No clear data is available to demonstrate that station staff training has improved expedition or clarity ©f communication in emergency circumstances. Following this tragic incident at Stratford station, the RAIB’s investigation included a learning point about communications, specifically that staff are reminded of the importance of safety critical communications when reporting and responding to incidents. This includes: ° accurately describing what is known about the incident
* avoiding using jargon 8

B providing accurate details regarding the location of the event ■ repeating back information to ensure a clear understanding is reached 0 ensuring that all parties involved know what is expected of them (paragraph 129). We undertake routine monitoring of safety critical communications which includes recordings of safety critical communications used during day-to-day activities as well as during incident scenarios. We monitor the communications from both routine and incident situations to make sure that the appropriate communications protocols are used during both. We assess our station colleagues’ understanding of effective safety critical communication. We have reviewed the competence assessment data for our stations’ operational staff and the quality of assessments are consistently high. Across the questions that they are asked on emergency communications 85 - 95% of candidates selected the correct answers. This indicates that our station colleagues have a strong understanding of the core safety critical communication actions required of them. During 2026, we will build on this by putting a robust evaluation framework that measures whether learning has taken place and whether knowledge and skills have transferred into the workplace for both train operators and station staff. The framework will be in place by December 2026. Conclusion I would like to offer again my heartfelt sympathy and condolences to Mr Mitchell’s family and friends. We are determined as an organisation to continue to learn lessons from this tragic incident. I hope this response is helpful and welcome. Please contact me if I can be of any further assistance.