[Guest Post by Jaroslav Valuch, Head of Campaign against Racism and Hatecrimes at Agency for Social Inclusion, The Office of the Government of the Czech Republic, Field Coordinator with the Ushahidi Haiti project, co-founder of the Standby Task Force and Ushahidi community leader. Jaro coordinated the Krizova Mapa Ceska for the 2013 Floods.]
[caption id="attachment_13664" align="aligncenter" width="500"]
Feeding Crisismap crowdsourced citizen reports to TV broadcast via Krizova Mapa Ceska[/caption]
The Crisis map of Czech republic had been deployed during regional floods in May 2013 by the Czech public television. It brought some practical lessons on the use of social media to crowdsource crisis data from the citizens as well as on the use of traditional TV broadcasting to feed curated citizen reports back to public. All of it in a mid-size country with standard emergency response mechanisms in place where disasters usually don't cause the whole system to collapse.
Having received a feedback that such project did help to save property, we had an unique opportunity to test the hypothesis of the need to close one of the gaps in emergency reporting and information sharing.
Before we look at recent emergency, let's have a look further back to past. It is summer of 2012 and the Czech television's (public broadcaster) new media division team meets with official responders and humanitarian organizations for the first time. The aim is to explore the possibility of collaboration on crisismapping project where public plays critical role in reporting. Being a co-founder of Standby Taskforce, I did the introduction of a concept of volunteer-powered live mapping and information processing, Pavlina Kvapilova, head of New media division of Czech television introduced their use of Facebook to crowdsource reports during floods in 2010.
The feedback was pretty calm and negative, something along: "Czech republic is not Haiti, we have systems in place and we know what we do and we do it right. Map of this kind will just add unreliable noise, spread hoaxes and will support volunteers to show up in uncoordinated manner. As responders simply put it: "The second disaster that comes after the first one are volunteers."
This was no shock for us. I am beyond the "tech can save everything" mindset already. The key was to identify the added value the project could potentially have for improved situational awareness and citizen response coordination. We simply believe that there is a great need for a communication channel where citizen crisis reporting can be leveraged to fill some cracks in the system and brings additional value through information curation process.
First of all, being the map an independent initiative, it would most likely end up with a map with some red dots. Period. The key aspect here is the fact that it was run by national public TV broadcaster and the reporting loop was closed.
Phase one clearly served for better situational awareness. Crowdsourced and curated data were fed back to public via special emergency broadcasting service.
In phase two once the immediate flood emergency receded we worked with humanitarian responders to assist in volunteers coordination. We displayed actual needs of humanitarian organizations as well as locations of volunteer coordination centers on the map. Volunteers were directed to map via broadcast to first find out if and where their help is needed.
It became obvious that people simply enjoy being reporters, providing unique audiovisual footage for the broadcast and consequently to their fellow citizens.
We were pleasantly surprised by the quality reporting. Reports were factual and supplemented by pictures and videos.
[caption id="attachment_13662" align="aligncenter" width="482"]
Email quote from women whose property was saved thanks to unverified citizen report of flood emergency that wasn't reported by official sources on time[/caption]
The news studio enjoyed having fresh audiovisual content, not having to keep recycling the same footage all over again.
In early days we didn't manage the servers not to crash after every mention of the map in TV broadcast. Inaccessibility of the map was the main reason for negative feedback.
The key finding is that during follow up publicity, nobody disputed the principle of citizen crowdsourcing itself. It is the duty of a public broadcaster to provide crisis reporting services in emergency, and via crisis map it extended the notion of this role to a new level.
We didn't use mobile app, simply to be able to process reports submitted via web and Facebook. Next time we go with app thanks to increased team processing capacity.
Twitter turned out to be least useful source given its low popularity in country, useless content and number of retweets.
We had verification processes in place. However, they were used only for potentially sensitive reports. In general, only reports that could cause serious panic were verified via phone with police or local municipalities.
Furthermore, the crowdsourcing of verification evolved naturally as people themselves started correcting some of our reports and providing additional info and audiovisual evidence. For the future, we plan to systematize this.
[caption id="attachment_13661" align="aligncenter" width="500"]
Crowdsourced verification - Crowdsourced verification - citizen corrects original unverified report of water breaking into city, comment of the CM team added stating that based on this correction the situation was checked with local authorities[/caption]
Many local municipalities were failing to provide up-to-date crisis information through alternative channels. If their capacity increased, it would enormously help our verification team.
Positive feedback is based on anecdotal evidence, we still need to do proper evaluation.
Crisis map volunteers managed to jump on board quickly, without any prior intense training.
Complaints we received during deployment mostly focused on the fact that only the capital city Prague was being extensively covered while remote areas were not. We shifted our focus to remote areas, and got complaints that we don't cover Prague. Anyway, good news for the team as it proved that people demand such kind of distributed reporting.
Primary purpose of the crisismap was not to improve the situational awareness of official responders ("we have all info we need from our sources"). Follow up response from the central emergency coordination body was that they didn't find any use for our map. However, we were approached by official regional emergency coordinators who identified a great potential use of crisismap for their local operations during emergencies. Planning of joint simulation is already underway.
Feeding Crisismap crowdsourced citizen reports to TV broadcast via Krizova Mapa Ceska[/caption]
The Crisis map of Czech republic had been deployed during regional floods in May 2013 by the Czech public television. It brought some practical lessons on the use of social media to crowdsource crisis data from the citizens as well as on the use of traditional TV broadcasting to feed curated citizen reports back to public. All of it in a mid-size country with standard emergency response mechanisms in place where disasters usually don't cause the whole system to collapse.
Having received a feedback that such project did help to save property, we had an unique opportunity to test the hypothesis of the need to close one of the gaps in emergency reporting and information sharing.
Before we look at recent emergency, let's have a look further back to past. It is summer of 2012 and the Czech television's (public broadcaster) new media division team meets with official responders and humanitarian organizations for the first time. The aim is to explore the possibility of collaboration on crisismapping project where public plays critical role in reporting. Being a co-founder of Standby Taskforce, I did the introduction of a concept of volunteer-powered live mapping and information processing, Pavlina Kvapilova, head of New media division of Czech television introduced their use of Facebook to crowdsource reports during floods in 2010.
The feedback was pretty calm and negative, something along: "Czech republic is not Haiti, we have systems in place and we know what we do and we do it right. Map of this kind will just add unreliable noise, spread hoaxes and will support volunteers to show up in uncoordinated manner. As responders simply put it: "The second disaster that comes after the first one are volunteers."
This was no shock for us. I am beyond the "tech can save everything" mindset already. The key was to identify the added value the project could potentially have for improved situational awareness and citizen response coordination. We simply believe that there is a great need for a communication channel where citizen crisis reporting can be leveraged to fill some cracks in the system and brings additional value through information curation process.
So what were the biggest concerns humanitarians had?
- The map and social media will only add noise and unreliable information to already confusing online space
- Peoples' expectations will rise once we encourage them to report in crisis - having a chance to report your needs, you will expect someone to come and meet those needs.
- Volunteers will use the map to find where the help is needed and will take action that will only cause logistical problems to official responders and municipalities
So how did it turn out finally? See the list of our findings.
Info box: Recent floods in Czech Republic caused death of 15 people, 26 thousands were evacuated, 970 municipalities were affected, damage reaches 10 mil. USD During ten days of Crisismap deployment, 378 000 page visits, 2.5 million pages displayed, 3000 reports collected via map and 4500 via Facebook page, almost 2000 repots finally made public on the map. 14 volunteers joined deployment, 5 members of new media division participated.
First of all, being the map an independent initiative, it would most likely end up with a map with some red dots. Period. The key aspect here is the fact that it was run by national public TV broadcaster and the reporting loop was closed.
Phase one clearly served for better situational awareness. Crowdsourced and curated data were fed back to public via special emergency broadcasting service.
In phase two once the immediate flood emergency receded we worked with humanitarian responders to assist in volunteers coordination. We displayed actual needs of humanitarian organizations as well as locations of volunteer coordination centers on the map. Volunteers were directed to map via broadcast to first find out if and where their help is needed.
It became obvious that people simply enjoy being reporters, providing unique audiovisual footage for the broadcast and consequently to their fellow citizens.
We were pleasantly surprised by the quality reporting. Reports were factual and supplemented by pictures and videos.
[caption id="attachment_13662" align="aligncenter" width="482"] Email quote from women whose property was saved thanks to unverified citizen report of flood emergency that wasn't reported by official sources on time[/caption]
The news studio enjoyed having fresh audiovisual content, not having to keep recycling the same footage all over again.
In early days we didn't manage the servers not to crash after every mention of the map in TV broadcast. Inaccessibility of the map was the main reason for negative feedback.
The key finding is that during follow up publicity, nobody disputed the principle of citizen crowdsourcing itself. It is the duty of a public broadcaster to provide crisis reporting services in emergency, and via crisis map it extended the notion of this role to a new level.
We didn't use mobile app, simply to be able to process reports submitted via web and Facebook. Next time we go with app thanks to increased team processing capacity.
Twitter turned out to be least useful source given its low popularity in country, useless content and number of retweets.
We had verification processes in place. However, they were used only for potentially sensitive reports. In general, only reports that could cause serious panic were verified via phone with police or local municipalities.
Furthermore, the crowdsourcing of verification evolved naturally as people themselves started correcting some of our reports and providing additional info and audiovisual evidence. For the future, we plan to systematize this.
[caption id="attachment_13661" align="aligncenter" width="500"] Crowdsourced verification - Crowdsourced verification - citizen corrects original unverified report of water breaking into city, comment of the CM team added stating that based on this correction the situation was checked with local authorities[/caption]
Many local municipalities were failing to provide up-to-date crisis information through alternative channels. If their capacity increased, it would enormously help our verification team.
Positive feedback is based on anecdotal evidence, we still need to do proper evaluation.
Crisis map volunteers managed to jump on board quickly, without any prior intense training.
Complaints we received during deployment mostly focused on the fact that only the capital city Prague was being extensively covered while remote areas were not. We shifted our focus to remote areas, and got complaints that we don't cover Prague. Anyway, good news for the team as it proved that people demand such kind of distributed reporting.
Primary purpose of the crisismap was not to improve the situational awareness of official responders ("we have all info we need from our sources"). Follow up response from the central emergency coordination body was that they didn't find any use for our map. However, we were approached by official regional emergency coordinators who identified a great potential use of crisismap for their local operations during emergencies. Planning of joint simulation is already underway.
So, what's next?
- Evaluation with responders and identification of further improvements
- Improved list of categories that more closely respects what people spontaneously report rather than what we think they should report.
- Integration of more map layers and networking with number of new initiatives, such as I have/ I need applications that emerged.
- Strengthening verification procedures and crowdsourcing of verification
- Stable servers
- Mobile reporting