Tuesday, June 06, 2006

Web based open source disaster management system for Indonesian earthquake

Over the last few days I have helped arrange for the deployment of a web based open source disaster management system for the May Indonesian earthquake. An Indonesian rescue group will use the system, with technical support from Indonesian IT university students and funding from the Australian Computer Society.

This came about when a few weeks ago Ken Taylor from CSIRO asked me if there was a project I might have for some students. They are at an Indonesian university studying IT. Most likely they will get work doing software for another country so needed practice in "offshoring". Also earning a little money for themselves and the university would be good.

Mount Merapi in Indonesia was threatening to erupt, so I suggested they could look at if Sahana was suitable for deployment. That way we might have a system ready when there was an eruption.

Sahana is open source software originally developed for the Asian tsunami but now generalized for any disaster. It is designed to help run refugee camps. I helped with some of the user interface, to make it more efficient and see if it would work on wireless devices. I gave talks in Sydney at an IT conference and to government webmasters in Canberra to see if it could be deployed in Australia for a bird flu pandemic, but Australians would be reluctant to use something which was free (and therefore suspect) and also was developed in the third world.

So I arranged for the ACS to fund the students to see if Sahana could be used for Indonesia. Rather than a traditional secret report, the students used a WiKi to write their report, allowing anyone to edit it.

When the report was just about finished I had them copy bits from the WiKi to a slide show and I presented that to the ACS Council's national meeting in Canberra. The same day I presented, the Indonesian earthquake struck.

The students had reported the system was suitable for use in an Indonesian disaster. We now had a disaster, so the obvious thing to do was deploy the system. The students already had the system running, the ACS could provide more support, but we needed a customer.

What few people realize is how difficult it is to contact anyone in authority during an emergency. It is even difficult to work out who is in authority, or for them to know they have authority. If you find someone, it is difficult to then suggest they do something they have not planned. During an emergency staff are busy carrying out their planned tasks, using rehearsed scenarios. They last thing they want to do is listen to long presentations from someone they have never heard of, about some computer system.

Under normal circumstances our next step would have been to demonstrate the system to government and non-government authorities so it could be included in their plans. I realised this would not work. What we needed was a credible aid group willing to try the system. But there was not time for formal approaches. At ANU we happened to have an Indonesian staff member, Dr Teddy Mantoro, who knew people in the Indonesia Whitewater Association. They had done some rescue work and had people already assisting with the emergency.

I was initially skeptical about this as it would not look good to be helping a recreational sporting group. But when I checked I found they were trained in rescue in Australia, by an Australian government accredited organization. So we got them to ask for help.

The plan which evolved was for the students to run the system and IWA use it. I had the students clone their existing web page to produce one to coordinate the project. Open Source and emergency experts around thew world are invited to help by editing the WiKi.

One problem I had to overcome was my reluctance to edit the WiKi. What I wanted to do was tell someone else to do it. But there isn't time to do that. The working approach therefore is to make a change and record why you did it.

Suggesting what should have been done, after a disaster is one thing, but deciding what to do during one is an altogether more difficult task. This is a very different process to commenting on 911, the Canberra bushfires or even help with military exercises, as I have done in the past.

2 comments:

  1. See article about it "Third World Software Could Have Role in Disaster Relief" by Sue Bushell, CIO Magazine, 08/08/2006 10:28:30 http://www.cio.com.au/index.php/id;464014852;fp;4;fpid;21.

    ReplyDelete
  2. Version of this Blog posting in The Australian newspaper, 27 June 2006: "Technology lifeline in volcanic zone".

    ReplyDelete