Saturday, 15 February 2014

Emergency Management's collision with Social Media

Prime Minister Cameron tweeting where to find latest weather information

Social media tools during a crisis or disaster has become a necessity in many regions of the world. It has proven to be a resilient communications suite of platforms capable of reaching directly to the public when other forms of information distribution are impaired or not available. 

It works, but...

The most popular applications to use are Twitter and Facebook. Others like YouTube, Instagram, Crisis and Disaster Management Community - Education & Training are useful for post disaster information events. Twitter is easy to use, broadcasting instantly the moment a tweet is sent. Facebook is just as fast, allowing more detail to be published, not being limited to only 140 characters or single image per notice published. But there are a number of concerns and challenges when using social media including information confusion, multiple sources, and accuracy. Social media is considered by many to be a one way conversation and limited effectiveness due to the lack of resources and accountability that can be attached. Of even more concern is the scale in which social media operates. The environment is wide open and littered with abuse and misinformation. As events unfold, millions of tweets and Facebook posts explode on the internet. Coping with all this data and information are the Emergency Management Offices that are not equipped with hundreds of staff to analyze the data. Automating and filtering the data are effective tools that can qualify and pinpoint valuable information. It will mean significant upgrades in process, training, and skills will be needed. The value of social media's links between public safety and the public currently faces not only these challenges but also acceptance on both sides of the problem. There are generation gaps, adaptability, and capital costs government agencies question as to its effectiveness.

Bridging the digital divide

There can be no question as to social medias impact on disaster response. It is a game changer in how the public and government civic protection agency's deliver information. But how best use social media is a question with no single answer or solution. The most common approach is to use social media as an announcement platform, ensuring warnings and what government actions are being undertaken are distributed to the public. It offers multiple paths of redistribution as well, as most if not all mainstream media organizations use social media as a source of stories and updates. In some countries social media is the only platform left to get information out in any form because it can be delivered from any origination point in the world. But of significant concern is the question, is anybody listening? As illustrated in the picture above, Prime Minister David Cameron's Twitter account, tweets where to find the latest meteorology reports and who to follow. The PM's Twitter account has 2.57 million followers. A pretty good subscription rate for a government leader. But how many of those 2.5 million followers are actually in the United Kingdom, or in areas where people are affected? It is certainly not the PM himself doing the tweeting either, but a senior communications officer assigned to the PM office, or at least reviewed prior to release. In fact, very few political leaders or agency leads actually run their own Twitter accounts.

Do you follow?

In the Philippines during Super Typhoon Haiyan - Yolanda, several lead agencies such as the Department of Social Welfare and Development (@DSWD, the National Disaster Risk Reduction & Management Council (@NDRRMC) were sending out thousands of tweets to notify the public. DSWD has 6,500 followers and tweeted over 3,500 times during the Typhoon. The NDRRMC has 64,000 followers in an area that has 2.5 million people affected. Many would argue that its use during the disaster was not an effective distribution mechanism. Not only was its distribution (and redistribution) hampered by lack of followers, but the ability to received social media was severely limited across the region for over three weeks in the most heavily damaged areas because all forms of communications were offline. Social media services were used as communications services came back online. The value to the local public was very little at that stage of the disaster. The one area social media was of immense value, was telling the world what had occurred and what supplies and needs the region needed. The world was soon flooded with pleas for donations, supplies and facts on what occurred. Millions of tweets were sent and retweeted by millions more. NGO's and international response agencies responded in kind, telling the world how it was going to respond and deliver relief. Local, regional and federal government agencies of the Philippines had nothing to do with these post event interactions and nor did they assign dedicated resources to these actions. They did not need to. Some would argue this was a wise decision. Federal government agencies did use social media to quell false rumors immediately after the disaster, such as the concern import taxes and stated delays in getting relief supplies moving once they arrived. But this was to notify the world and not the local population.

Tweets are not mechanisms of P2P direct action

Multiple government agencies in the United Kingdom are essentially using social media as part of their information distribution policy in the same way the Philippines did during Typhoon Haiyan - Yolanda. Social media is an excellent medium to let the country and the world know what governments are doing during a crisis. What governments are not prepared to do, is rely upon its use by the public in calling for help. This is particularly true during large scale disasters where there are hundreds of thousands affected. Local, regional and federal agencies are not staffed or equipped to handle it, along with supporting traditional mediums of emergency response tools. This is not true in all parts of the world, but generally holds true. There are real accountability and scale issues that cannot easily be sorted out by software programming or automated analysis. It's not a question of can it be done, it can. Current rules and regulations are not easily adaptable to social media and internet technologies. Even with updated protocols, the most significant obstacle is ensuring the public knows what they are and building trust. Currently, the amount of false - positive data points are not accurate or sufficiently accountable to satisfy government or the public. There are a variety of proposals and standards on the table such as Social Media Emergency Management (SMEM), Humanitarian Exchange Language (HXL), #Hashtag Protocol, and many more. None are universally accepted. Several countries do have regulated information specifications to create fully integrated standards within government agencies. The National Incident Management System (NIMS) has proven effective since its creation by FEMA. It went through significant obstacles and hurdles before its adoption was widely accepted. Some non-government organizations are integrated into NIMS such as the American Red Cross, but most are not. NIMS was never designed for non-government systems of information coordination or support. 

There is a need to develop one in the social media world. I estimate that during a disaster, over 80 percent of tweets sent are ineffective and of no value or impact or influence at the Emergency Management response level. In some cases, this is probably closer to 95 percent. What worries leadership are the remaining 5 to 20 percent. One example was during Typhoon Haiyan - Yolanda, during a time of significant confusion regarding the delivery of aid to Romblon - was it delivered or not to the region, which was in desperate need, 14 days after the Typhoon had hit. On November 22nd, I tweeted to the NDRRMC, asking for confirmation of delivery. Amazingly, I got a reply within an hour. It was not an answer we wanted to hear, but a confirmation that they did not know themselves. NDRRMC transparency was refreshing to witness. The workload at the time was overwhelming and they still managed to reply. If this scenario had played out in a densely populated city in Canada, U.K. or the U.S., I doubt a reply or acknowledgement would have occurred at all. The agencies are simply not equipped to handled the volume or have detailed information asked at their fingertips and not respond. Social media is not a emergency management duplex environment readily adaptable to existing systems.

Integration and Syllabus

Social media is more than just tweets or Facebook postings. It is the collaboration and awareness potential that offers response agencies and the public, effective survival and recovery information before and after a major event. This is where another collision occurs, information overload along with identifying pertinent information in the same consistent manner, enabling instant recognition of what the information means. Currently we have multiple sources of Geo-spatial Information System (GIS) published maps with preexisting data points that are easily augmented in updated metadata. But there are no current standards being followed, resulting in multiple interpretations of what information is being posted. There isn't even a standardize usage of public safety symbols being used across all map publishers or authors of updated metadata. Presently, every map and social media inflow of metadata published is all ad-hoc, with no one single entity responsible for adherence or compliance to a standard. We have road signs that are universal around the world, yet we can't seem to collectively agree on symbols that should be used on all maps. GIS powered maps are powerful tools used by various disaster response stakeholders. Many of agencies want to integrate and fuse data they have into a coordinated and situation aware platform. This is beginning to occur and is touted as the next advancement in disaster response technology. We are already seeing adoption and integration of Open Data Sets into this environment. The next step is to layer social media metadata that is qualified and adds value. There are hundreds of illustrated examples where this has been successfully merged and published during a real disaster (Christchurch, New Zealand, Australian Floods of 2010, Haiti Earthquake, etc.) that are the potential stepping stones leading to a uniformed and calculated synergy of GIS and social media metadata.

Way forward

There are signs of improvement, both in the responsible use of social media and its ability to make a difference in civic protection. During Hurricane Sandy, the supervisor of New York City's Emergency Management Twitter account was filtering tweets immediately after it hit the city for hours during each shift and had come across a tweet that a family of four was in serious trouble and needed help. Through this individual's direct action, emergency response units found the family and rescued them on Staten Island. (you can watch a video of this event in the Crisis and Disaster Response Magazine in the documentary on the disaster of Super Storm Sandy) The reality is, if you have a data connection, you also have access to Voice over I.P. enabling a user to call a Emergency call center, but they are often out of service or busy. Using Twitter is a valuable option available to Civil Protection agencies. It gets a user into the Que enabling a response, if someone is listening. Like email, it could go unanswered for days or worse, never seen if filtered as spam either by the provider or the receiver. 

It is easy for all of us to suggest that governments institute and require the integration the monitoring of Twitter, Facebook, YouTube and other forms of social media into Emergency Dispatch Call centers and you fixed the problem. The short answer is, there is a very long way to go before such as capability can be even considered, let alone attempted. It has taken years for Emergency Operations call centers to accept and adapt mobile phone based calls into their systems. Social media may not require the same amount of time, but I would suggest, many agencies will take the position of not enabling it. This is because due to costs in training and monitoring required, as it is a significant burden that has little proof that such investments are cost effective. This is not to say it cannot be effective in many parts of the world where very little legacy infrastructure exists. Social media integration in these environments might be the ONLY solution available for them that makes sense. In regions like Africa, the Middle East and sparsely populated areas in western China, governments could find Social Media technology is the only appropriate system and method to deliver public safety services.

I do not think we will ever reach global universal guidelines or social media standards used in crisis and disaster management. What we will see is increased use and implementation of social media products and services uniquely implemented based on local and regional needs. We will see more and more data available and applications developed to enhance response capabilities because of social media's popularity, but there will be no one size fits all.  The collision has occurred during a time where full coverage insurance is not yet available.


No comments:

Post a Comment