Communicating in emergencies is part of managing emergencies

Man in waterproofs talking to a camera held by another man in waterproofsI have a well-rehearsed line whenever anyone suggests I shell out good money to attend a conference

“I no longer attend conferences unless I’m speaking at them”

This is mostly true. I’ve spent too many days spent listening to disappointing speakers who seem only to be there because they paid for the honour or because the conference organiser doesn’t actually understand what the state of the art is. Combine that with the discovery of different ways of meeting and learning (like unconferences) and it is a very rare programme of speaker after speaker that will get me to travel anywhere.

But the Emergency Planning Society Wales Branch proves that there is still a place for a bunch of people in a room watching some powerpoints.

I don’t think it’s rocket science. It’s an event for a very clear (if small) audience put together by people who understand that audience and really care.

Plus emergency planners in Wales seem to be a universally pleasant group of people.

Anyway we had a tour through evacuating students, the emergency planning aspects of taking action on modern slavery (that’s right, there are), considering the long term impacts on individuals affected by emergencies, interruptions to gas supply and much, much more.

Throughout all the talks, the importance of effective communications was stressed again and again. Communicating with affected communities, with the wider public and with journalists and the media is something that is increasing in importance.

And there seems little doubt that emergency planning professionals recognise the need to have trained communication professionals involved throughout emergencies and recovery.

Which, it won’t surprise you to learn, is a position I wholeheartedly support.

But I am also concerned. Comms teams are under considerable pressure as the public sector makes deeper and deeper cuts. That’s life right now in the public sector and it has to be managed. The temptation to cut back on training and exercising comms staff, on reducing rotas, on ending on-call payments must be strong. Maybe these things could seem like a luxury in a time of austerity.

One sensation that all emergency planners recognise is the feeling when an incident starts. The uncertainty about how bad things will get, who will be affected, what will need to be done to keep them safe coupled with the certainty that no-one else is coming to help.

We can’t magic money out of nowhere but we must not forget that communications is a fundamental part of emergency response and requires skilled, trained, experienced professionals to deliver.

This might seem slightly self interested since I run a company focused on training comms officers to work effectively in emergencies. But really it’s the other way around, I think you need effective communications to effectively manage emergencies. And I think I can help.

Photo is Morpeth Floods by John Dal used under a Creative Commons licence

Things we should sort out about using open sources in emergencies

Numbered cogs and metal buttons on a machine in close up

A problem

There’s a problem in the use of open source intelligence in emergencies.

Or perhaps more accurately there are a series of ineffectively managed risks.


I think that, at least in prinicple, there is widespread acceptance amongst category one and two responders and others that digital communications tools play a key role, possibly the primary role, in warning and informing the public. That was not the case even a small number of years ago. The risk that someone senior would turn up in an emergency and say

“we’re not doing any of that Facebook nonsense”

has diminished. There remain considerable differences in culture and capacity around digital between different organisations and this does create an area of some risk.

That’s not the risk I want to talk about here though.

As organisations get more used to using these tools in emergencies they are starting to notice that there are useful data being shared in social networks. Increasingly managers are asking for information about what is happening online.

This is all to the good of course and one of the reasons I am backing the plan to bring the VOST concept to the UK.

It’s also where the risks come in.

Who is watching social media?

Often when you pose these questions in multi-agency groups a police officer with plenty of stripes will bristle slightly and point out that it is self-evident that the police service does the intelligence gathering around here.

That’s fine, police forces are clearly equipped for such work and I guess the average citizen would expect police forces to be gathering data from open sources around, for example, a controversial protest march. Are they as skilled in gathering data relevant to surface water flooding? Or animal diseases?

The risk is that if we aren’t clear on whose job it is, it becomes no-ones job.

Or the person who everyone looks to do it in an emergency gets made redundant.

You lot know about Twitter

And my experience  is that in many organisations communications (media / PR / digital) teams are being asked to play this sort of role. Sometimes explicitly, sometimes tacitly.

On one level, again, that’s fine. Comms teams usually have the technical skill and familiarity with social networks.

Comms teams are trained for a different task. Mining data from social networks (as I do in my voluntary role with Standby Task Force) is a skill. It requires judgements to be made based on a set of incomplete understandings. Just like other forms of intelligence gathering (or research).

I’m not saying comms professionals aren’t capable of doing this, but are they being trained to make sure they are analysing the data objectively and presenting reports with accurate confidence weighting?

Reasonable searches

And I’m not convinced we’ve got the issues around what level of data we should be mining sorted yet. Running a search for mentions of place names in case people are telling each other (but not the council) of flooding sounds reasonable (doesn’t it?).

What about when I see a report from an account I’m not familiar with, how do I work out whether I can trust that report. I’d probably have a look down their social media profile, to see if they have sent messages about that location before. I might Google their name (or their user name) to see if I can find out more about them on other social media platforms.

Is this reasonable and proportionate?

That citizen has a right to privacy but it’s not an absolute right. The state (including, presumably, local authority comms officers) can infringe people’s privacy if it is lawful, reasonable and proportionate to do so. And the fact that people have put information about themselves in places where it can be seen does not, of itself, mean it is reasonable for me to go and look for it, in this context, at this time.

These feel like risks we should be talking about. They are all highly manageable through the key tools of emergency management: planning, training and exercising.

It’s certainly something we hope to build into the VOST model for the UK.

It’s just ignoring them that is risky.

Podcast: second album syndrome

Against all expectation Helen Reynolds and I have managed to produce a second podcast. The sound quality isn’t perfect as we were wrestling with flaky hotel internet.

But we did manage to cover NATO, the joys of the Economist, a smart NHS marketing campaign, encryption and loads of other stuff.

Hope you enjoy it

If you really like it, why not subscribe on iTunes?

VOST in the UK: an actual plan

Day two at BlueLightCamp was a hack day.

But we encouraged everyone to think a bit beyond the usual conception of hackdays (lots of dev, lots of code, lots of visualisations) to focus on getting useful work done in any medium.

So a group of us foregathered to carry on the VOST discussion that had kicked off on day one and move it into an action plan.

A photo record of the flipcharts we threw out is available.

Got to start somewhere

We started on from the assumption that having VOST available in the UK would be a good thing, which is not to say that we believe the argument has been won (or even had) but you’ve got to start somewhere.

Then we imagined a hypothetical police officer chairing a tactical meeting. They get to the item in the checklist that says “consider activating VOST”.

What do they need to know to make that decision? we wondered and suggested these sort of things

A well understood set of terms of reference.

Realistically, the nature of the UK approach to emergency management means clarity on this point is very important. One model suggested to emulate is the casualty bureau. Police forces maintain this capability, make it available to other agencies in emergencies and the scope and capability is widely understood.

One distinct feature of VOST is it is agency neutral. That potential means no agency owns it but also means it can be nimble and responsive.

Is VOST appropriate for all incidents? Would you activate a VOST in a suspected terrorism incident? Would it be activated in a public health emergency?

Clarity on the outputs from VOST.

Is VOST a filter finding the items that look relevant or will it apply intelligence. If the latter, how will the certainty in each item be communicated. Will VOST provide written sitreps or map overlays or something else (live access to a spreadsheet for example). There was a lot of excitement around the idea that we could produce mapping layers with different colours or coding for different reports and confidence. This pleased me as it seemed we were reinventing Ushahidi.

How often will VOST report. In a presenting incident a “battle-rhythm” (I don’t like these military allusions myself) of hourly updates was considered a good target. Which raises the question of what level of resource is necessary to provide that with no notice for a couple of weeks. Then there is the recovery phase, where a slower rhythm is probably appropriate but that could go on for a long, long time.

Where it fits into multi-agency response.

Actually we went round this quite a bit. In terms of where cat one responders seem to be building their capacity VOST would link to comms / media cells. There are other attractions to this from my point of view… comms cells are multi-agency and tightly integrated into the tactical layer.

The argument against this is that VOST (certainly as we conceptualised it) is primarily an information gathering/processing function so that would seem to place it in the intelligence cell which would bring a whole set of complications.

Or there could be a distinct VOST cell. As I understand it this is essentially the model used by VOSTs in some other countries. Personally I’m quite attracted by this model but I think, practically, it would mean VOST would be left out more frequently.

And this led us into some slightly wider questions.

Is any LRF / RRP already doing this sort of thing well?

This seems like a good question for the CCS, or the internet generally.

What are the practical and operational limits of VOST?

And related to this what are the technical requirements. Personally I feel that porting the VOST workbook to Resilience Direct would be a great place to start but that would make it hard to collaborate with non UK VOSTs. Which raises the question…

What should be the link in terms of interoperability with VOST volunteers around the world?

How directly applicable is the global / US model to the UK?

This is a genuine question. On the face of it if it works in the USA you would have thought it would work in the UK but we do have a very different concept of emergency management and some very different organisational cultures. So we feel we need to test this

And then towards some actual proposals:

To develop a “product” that can be offered to LRFs to plug in at tactical level.

There are some prerequisites that we see to this some of which are listed above (sorting out the TOR, the capability, the limits and the outputs).

We need to nail the data protection and Article 8 (of the European Convention on Human Rights) issues too.

And we need to do something so we can show people what we are talking about.

To run a live exercise analysing data and producing outputs based around a non-emergency situation.

We’ve been calling this a “stress test”.

We started with the idea of Eurovision but we’d have to wait until next May to do that and we’d like to do something sooner. So we were thinking of a big cycling event or similar. We hope to recruit and train a bunch of volunteers just for that single event. This should help us test whether the USA model works in the UK and will give us something more tangible to show to LRFs.

We’d like to find a friendly academic who would like to provide some independent evaluation of this process. If you know anyone, point them our way.

Provide a competency framework and role profile for VOST team members.

There are potentially existing  frameworks and structures we could tap into for this. But we want to make it really easy for an LRF to get this capability up and working without having to do re-work.

We’re running a Trello board to coordinate this, it’s private for the moment but drop me a line if you’d like to be added in.

All comments, improvements, suggestions or, really, anything on this really welcomed.


Discussions on VOST at BlueLightCamp

BlueLightCamp took place in Birmingham last weekend (6 and 7 June). I expected there to be some discussion on VOST* and there was.

I wasn’t expecting there to be so much discussion, and planning, but there was.

@Rubonist pitched a session on VOST on Saturday and several of us who think the idea has merit for the UK got together with some who are unconvinced and some who had never heard of the idea.

I thought this was a really useful discussion. It crystallised some thinking for me, especially the fact that we need to separate out the way of working: the VOST mission as it were, from the staffing (volunteers or paid staff). We need to recognise that actually there are several conceptual leaps that people need to make if they are to get their heads around what we are proposing:

  • there is useful data in social networks
  • that data is not currently available to decision makers in emergencies in the UK
  • if it were available they would make better decisions
  • processing that data requires specially trained people able to work together
  • they don’t need to be in the same location
  • those people could be specially trained volunteers (or specially trained employees)

I think I tend to jump straight to the end without checking everyone is with me through all the stages.

What was particularly significant about this for me was that (@mtthwhgn‘s point) we can talk about VOST as a process within multi-agency response without getting bogged down in the question of volunteers. So we might say it is a task for the comms/media cell, train comms officers and make sure that they have the ability (and technology) to operate a VOST.

I suspect (as I imagine most comms professionals in cat ones will) that comms teams won’t have the capacity to do that on top of their warning and informing tasks.

Which is why we keep talking about volunteers.

Not all of those who started the discussion unconvinced by the need left it with their minds transformed but I do think most people felt that there was a potentially useful role here which would fill a gap to some extent.

I went to an earlier workshop pitched by Neil Beet about, essentially, open source intelligence gathering (in terms of legal and social limits). This is a particular interest of mine. I think it would be useful to nail this issue in terms of VOST sooner rather than later.

I felt pretty positive about the engagement on VOST after day one, but there was more, much more, to come on day two.

*Virtual Operational Support Team. The very good idea that you can improve your information picture if you get a team dedicated to reviewing relevant social media stuff and telling you what is happening out there. They do it in other countries already.