Michael Hallowes is a distinguished leader in emergency management, known for pioneering Australia’s ‘Emergency Alert Program’ and introducing performance standards in Victoria. With a 30-year career in British policing, including roles at New Scotland Yard, he concluded as the operational head of the UK Police Staff College. Michael has written this article for NPC in which he reflects on lessons learned from Australia and Canada in developing and implementing emergency alert systems, and the journey taken towards developing similar systems in the UK, with the challenges along the way.
The Civil Contingencies Act (Part 1) 2004 makes clear that ‘Category 1 Responders’ (the emergency services – predominantly police and fire) are “responsible for warning and informing the public in relation to emergencies”. Developing a national UK public alerting system to effect this statutory ‘duty to warn’ only began in 2013. That was when the Cabinet Office Civil Contingencies Secretariat (CCS) first conducted two localised tests using the available mobile phone technologies, (Cell Broadcast (CB) and Location-Based SMS (LB SMS)). CCS determined the latter provided greater functionality, but no further significant action followed until 2018, when CCS appointed Fujitsu consultants to help develop the project.
In July 2015, I came to London to give my first briefing to CCS to outline the successes with ‘Emergency Alert’ that I had led in Australia, and guide progress on CCS’s own project. Four years later, I ran a workshop for the Mayor’s Office for Policing and Crime (MOPAC) to develop the ‘Concept of Operations’ and ‘User Requirements’ for London’s emergency services. The objective was to help CCS accelerate its own project to finally implement a UK emergency alert system. Both the CCS project lead and their Fujitsu consultant participated. The workshop was instigated in response to specific recommendations in Lord Toby Harris’s independent review of ‘London’s Preparedness to Respond to a Major Terrorist Incident’ (October 2016), which stated, “The Mayor should quickly work with the Cabinet Office to introduce a London-wide pilot of public alert technology”.
The outputs from the workshop included a requirement for a granular, localised public warning system identical to that used in Australia. Forefront in our minds were the words of Michael Dowden, Watch Manager at North Kensington Fire Station, in his evidence to the Grenfell Tower Fire inquiry on 26th June 2018:
“For me to facilitate and change a stay-put policy to a full evacuation was impossible…I didn’t have the resource at the time. We’re looking at 20 floors above the fire floor with just six fire engines in attendance, one central staircase. I just don’t know how that could be done with the resources we had in attendance at that moment in time.”
In October 2018, I briefed the then Commissioner of London Fire Brigade and her executives on the advantages of a public warning system equivalent to Australia’s. They agreed that if Michael Dowden had been able to access such a system, firefighters could have saved more lives.
Every senior officer present at the MOPAC workshop wanted the emergency alert system to use LB SMS. This recognition stemmed from their acknowledgment of the numerous operational advantages derived from this technology’s ability to visualise data in near real-time. It provides precise insights into the exact number of devices or people within the warning area, showcases their concentrations through a heatmap, confirms successful delivery of the warning to the vast majority (95%+), and allows for tracking evacuations when the message dictates such action. They were in unanimous agreement that they did not want CB technology, because it had none of these advantages for the types of emergencies they envisaged using the system for, to help save lives. It is regrettable, that the Cabinet Office subsequently chose instead CB for the UK’s emergency alert system, given it has such limited utility for the police and fire services.
Learning lessons from Australia
The process followed in the UK differs significantly from my Australian experience. I was appointed as Emergency Services Commissioner for Victoria, Australia, in 2011. My Melbourne office was already focused on developing a location-based public alerting capability for Australia using mobile phones, also known as ‘Emergency Alert.’ The driver was ‘Recommendation 4.8’[1] in the Interim Report (August 2009) of the Victorian Bushfires Royal Commission. This public inquiry had examined the deaths of 173 people across the state during the ‘Black Saturday’ bushfires on 7th February 2009. The Commission recognised that communities need to be warned at the local level, via their mobile phones, based on their physical location in proximity to an imminent or actual threat to life. At the same time, three communities had begun a joint class action against the state government for a perceived failure by the emergency services to warn them of the approaching fires. In 2016, the Victorian Government finally agreed to contribute to the $794 million (AUD) paid in total by several parties in the out of court settlement. This should be a lesson for Cabinet Office on the potential legal liabilities should any UK police or fire service not be able to fulfil its statutory duty to warn through timely use of ‘Emergency Alert.’
In December 2012, we rolled out Australia’s ‘Location Based Solution’ which uses the LB SMS technology. We chose that over CB because it met the first functional requirement – that the system had to be accessible to and reach the vast majority of the public simultaneously, via their choice of mobile device on every network (which the rival CB technology did not). LB SMS also met all of the operational requirements determined by Incident Controllers in the emergency services. That is because the technology provides immediate access to the anonymised and aggregated location metadata generated automatically between each mobile handset and the network provider. The system presents this in near real-time on screen, to assure the Incident Controller that their message has been delivered successfully to at least 95% of devices located within the defined geographical area. The ‘Concept of Operations’ we developed at the outset was based on a number of use cases, including the need for localised alerts at the granular level, potentially, to a single high-rise building. On 14th December 2012, we demonstrated that capability by sending a test alert to the Department of Justice, a 37-storey office block in Melbourne’s Exhibition Street. The system delivered it successfully in less than ten seconds and limited the area to just that building and the 5,000+ people inside. This was very close to the scenario the Watch Manager, Michael Dowden, faced at Grenfell Tower four and a half years later on 14th June 2017.
Although developed initially for use in large-scale bushfires and floods, Australian emergency services rapidly recognised the value of public alerts for many other types of life-at-immediate-risk threats. About a month after the Lindt Café Siege in Sydney, in December 2014, I presented a paper to the Australia-New Zealand Counter Terrorism Committee recommending that we expand use of ‘Emergency Alert’ to all life-threatening incidents of any scale. That resulted in a policy which allows the police to use the system for high-risk crimes in action, such as an active shooter or a marauding terrorist attack.
By my return to the UK in late 2017, emergency services across Australia had used the system over 1,500 times to alert more than 15 million people to localised life-threatening events. Very often, the number of people warned was less than 5,000. No lives have since been lost due to a perceived failure by the emergency services to warn the public.
That remarkable success is largely attributable to the streamlined processes (from initial realisation by a first responder of the need to warn, through to an incident controller authorising use and an operator constructing and sending the alert) that reduced the activation time to a mere eight minutes. All actions and decisions are taken within the joint emergency services’ State Control Centre. Additionally, as part of an extensive programme of public education and operational preparedness, the system was tested in localised live ‘Community Based Trials’ multiple times during its development. This helped communities both get used to it and understand their responsibilities on receipt of a real warning. It also enabled rapid proficiency amongst frontline Incident Controllers and their staff on when and how to use the system.
Useability challenges of the UK system
The first public test of ‘Emergency Alert’ took place across the UK on 23rd April 2023. This exercise heralded the launch of the UK’s national public warning system after ten years of development. It was evident, however, that CCS had not learned from other countries like Australia, that had been using their highly effective system for eleven years without a problem.
The CCS test certainly failed Australia’s first requirement that ‘Emergency Alert’ “be accessible to and reach the vast majority simultaneously across all devices and networks.” As such tests are meant to do, it did identify significant areas for improvement. Cabinet Office remains silent, however, on what lessons it will take from other countries, and whether it will hold further tests to address current flaws. Notably, the first test demonstrated the need for major improvements before any police or fire service and the public can rely on the system.
What CCS has delivered is a public warning system best suited to an organisation, such as the Environment Agency, which has no responsibility for rescuing people, but is charged only with warning them of an impending danger. Had CCS accepted the London emergency services’ requirements for data visualisation, first responders would know how many people are affected, where they are and what resources they need to deploy in the rescue effort. SMS also enables two-way communication between the emergency services and the people affected both at the time of alerting and afterwards. Hopefully, Cabinet Office will learn from France and Australia, which have now implemented hybrid systems. These give incident commanders the choice of CB for the rare slow time-to-impact, wide-area emergencies, and LB SMS for the far more frequent fast-moving localised ones, where they need near real-time data visualisation from the ground.
In June 2022, I gave expert testimony to the Mass Casualty Commission public inquiry examining the murders of 22 people by an active shooter in Nova Scotia, Canada, over 18th-19th April 2020. I focused on the perceived failures of the Royal Canadian Mounted Police to use ‘Alert Ready,’ the Canadian equivalent of ‘Emergency Alert.’ Amongst my findings were:
1. Poor governance arrangements which led to the police being excluded from developing and updating operational requirements.
2. The project delivery team had not developed a Concept of Operations to define when and how the system would be used.
3. Frontline police officers had not been trained in when and how to use the system for this type of emergency.
4. Police did not have immediate, direct access to the system for sending timely warnings.
The outcome was that the Commissioners called for a “national review of public alerting.” I have now asked Cabinet Office what action they will take to review the UK’s arrangements for public alerting, as they appear to mirror my findings of Canada’s position.
There are other concerns about how the system will be used. On 28th June 2023 the Cabinet Office Minister, Jeremy Quin MP, responded to a written question saying:
“For an emergency alert to be issued, a major incident must pose a risk to life. Risks to health or property will not be considered reasonable use. Members of the resilience community who may need to request an alert have been informed of this threshold. Releasing an alert will require a request from the lead government department to the Cabinet Office National Situation Centre watchkeeper, and then approval by a Cabinet Office Duty Director. Where an alert is likely to be received by 1 million people or more, it must also receive ministerial approval. This system ensures that the Cabinet Office will only send emergency alerts when there is an imminent threat to life.”
It is apparent that this overly bureaucratic authorisation process will not be able to keep up with the dynamics of a fast-moving life-threatening emergency, such as we saw during both the Cumbria Shootings in June 2010, and the London Bridge terrorist attack in November 2019, let alone another Grenfell Tower-type fire. It certainly does not match the learning from Australia.
On the subject of when the public should expect an alert, the Cabinet Office website states:
“You may get alerts about:
- severe flooding
- fires
- extreme weather
Emergency alerts will only be sent by:
- the emergency services
- government departments, agencies and public bodies that deal with emergencies.”
Interested in their approach, I wrote to Cabinet Office using a Freedom of Information request to ask the following questions:
1. What police-related types of incidents would justify use of ‘Emergency Alert’?
2. How many police officers and staff have been trained and exercised in how to use the system?
3. How many supervisors and managers have been trained and exercised in authorising use of the system?
Their response stated:
“The protocol for sending an emergency alert will depend on the type of emergency and the scale of the audience. All emergency alerts must be authorised and sent by the COBR Unit, but they will be requested by emergency responders through Strategic Coordination Groups.
Whilst we can confirm that training is being provided to emergency responders on emergency alerts, we are unable to provide more detailed information due to national security rationale.”
This is not a very reassuring response and suggests that there is still a lot more work to be done by the Cabinet Office to develop the processes to ensure that the system can be used speedily and flexibly, to handle local-level as well as national emergencies. There is a wealth of international best practice they can draw on – most notably from Australia.
We all expect that in a life-threatening crisis, our local emergency services will have the tools to alert us quickly, explaining what we must do to stay safe. It is not clear that the UK Government’s cautious approach to emergency alerting can as yet provide us with that assurance.
Footnotes:
[1] The Australian Government, Council of Australian Governments and the State determine whether it is technically possible to implement the second phase of the national telephony-based warning system (that is, the delivery of warning messages to mobile phones based on the physical location of a handset at the time of the emergency) with a view to implementation for the 2009–10 bushfire season.