Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
name: Chamindra de Silva email: chamindra@opensource.lk R&D Manager, Virtusa (www.virtusa.com) Acting Director, Lanka Software Foundation (www.opensource.lk)
Table of Contents The Historic Trigger..............................................................................................................3 Why is the Response to Large Scale Disasters so Chaotic?..............................................3 The Need for Information and Information Technology in a Disaster..................................3 Introducing the Sahana Disaster Management System.........................................................4 Alignment to Free and Open Source Software (FOSS)........................................................8 Humanitarian-FOSS Concept and Awards............................................................................9 The Sahana Core Team and Community.............................................................................10 Application of Sahana in Sri Lanka, Post-Asian Tsunami 2004.........................................10 Application of Sahana in Pakistan, Post-Asian Quake 2005...............................................12 Lessons Learned from the Asian Tsunami and Asian Quake ............................................12 Recognition of Sahana and Humanitarian-FOSS................................................................13 Sahana Phase II Status.........................................................................................................14 Sahana Websites and Documentation.................................................................................16 Sahana Future Plan .............................................................................................................16 Other References:................................................................................................................18
the mother wailing and desperately searching until she finds the where abouts of her missing child; think about orphaned children traumatized waiting for a familiar face of their extended family to find them; think about the people starving and needing aid because their whereabouts have got lost in the chaos; think about camps waiting for the right medical supplies to treat the people within; think about the relief coordinators who have to make guesses as to where the limited relief goes and in what quantity. The unseen devil in these situations is in the scale of operation and in being able to account for each an every individual equally from their medical needs to reunification with family to their relief supply, which amounts to a whole lot of data. That is where information technology helps manage information. Through IT the right data can be shared and accessed instantaneously by gov offices, field operatives, the civil society, victims acquaintances and the victims themselves to enhance the relief effort. Ironically the government especially in developed nations often have in-house solutions that could handle certain aspects of the disaster management, such as tracking people. Unfortunately those systems and the data they contain are often protected and by policy cannot be shared with external groups such as NGOs or any ol' volunteer who wants to help, thus it ceases to be the consolidated solution for everyone. Additionally the government owned, command and control procedures and systems in such an instance can sometimes get in the way (due to the lack of resources on the ground) and it can prevent independent relief groups and volunteers from directly and immediately helping the victims. Yet at the same time everyone needs to be coordinated well and information needs to be shared to allow all relief entities to operate as one, such that the aid and services can be distributed effectively. This is where a centralized collaboration portal for all relief groups (including the government) provides a lot of value. And that is where information technology plays an important role, in the ability to manage the scale of information and to improve accessibility to that information to all groups. Ultimately we want to make sure every man, women and child is being accounted for and the relief to everyone is swift and timely. And for all I know during the Tsunami and Pakistan quake no complete solutions were put available to address a lot of these coordination problems.
Solution 1: Sahana Missing Person Registry The Missing person registry is an online bulletin board of missing and found people. It not only captures information about the people missing and found, but the information of the person seeking them is also captured, which adds to the chances of people finding each other. For example if two members of the a family unit is looking for the head of the family, we can use this data at least to connect those two family members.
Features include:
Meta data around the individual such identity numbers, visual appearance, last seen location, status, etc Sounds-like name search (using metafore and soundex algorithms) Uploading of a persons picture Grouping by family unit or other groups types
Problem 2: Coordinating All Aid Groups and Helping Them to Operate Effectively As One.
For the Tsunami we had a massive outpouring of support from INGOs, NGOs and the general civil society setup to help the victims. In Sri Lanka I believe we had about 300+ NGOs registered providing support. If all groups are not coordinated effectively it results in problems such as clogged up supply routes, competition for providing support in some areas whilst other areas suffer a dearth of support, double vaccinations and mis expectations. As a result all that goodwill and aid pledged will go wasted and under-utilized. However this can be an overwhelm coordination task for authorized emergency controller to do manually. An IT solution can help here where you have an organization registry where we keep track of who is doing what where and more importantly where nothing is being done at all (or there is no coverage of a certain service). This way they could even self-distribute themselves evenly across affected region just by being aware what other relief groups are doing.
Features include:
Capturing a comprehensive list of meta data on an relief organization and all the activities they have in the region Registering ad-hoc volunteers willing to contribute Capturing the essential services each group is providing and where Reporting on the converge of services and support in the region and more importantly where there are no aid services being provided
people in them. It also provides a GIS view to plot the location of the camps in the affected area.
Tracking of basic meta-data on the location, facilities and size of the camp Integration of google maps to provide a GIS view of the affected region Ability to customize the list of important facilities needed at a camp Basic reports on the camps and drill-down by region
Features include:
Basic meta data on the request and pledges such as the category, the units, contact details and the status Customizable category of aid Filtered search of aid pledges and requests Ability to track partial fulfillment of the request
Reference:
http://www.linux.lk/~chamindra/docs/Sahana-Brochure.pdf
Very few countries and organizations today can afford to invest a lot of resources in disaster management when there is no disaster present. While this is obviously true of poor, developing nations, it is also true of richer, developed countries as well because there are always higher priority items that need the funding. A FOSS approach provides a low budget, volunteer supplement and global way to build such systems There is not much commercial interest in developing solutions in this domain as often during humanitarian disasters software licenses are given freely and it almost seems unethical to restrict software. With FOSS there need not even be any delays in getting permission for a license as anyone as the
Also such system should be shared, developed and owned globally as the problems they address are all too common for any country dealing with a disaster. Effectively It should become a global public good. The FOSS development and community mechanisms have a proven track record to build such goods. The global community of IT volunteers who can contribute their goodwill to such causes by using their skills to develop and customize FOSS software for the disaster situations As in conflict situations, during disasters segregation arises between Gov and NGOs, NGOs and INGOs and worse relief groups and people. The main reason I believe is the urgent circumstances in this instance. So an open and transparent and globally owned system is more likely to be trusted to mediate between the groups Finally no two disasters are alike. There are often localizations and customizations needed for the software before before it can get applied effectively to the disasters. Some of these localizations include adding additional meta-data around the entities in the system or translating the system to handle entry in a particular language. With FOSS, the code is available for anyone to quickly pickup and make the needful customizations without restriction.
Going the open source way can address the above concerns and using the open source development model, it is possible to develop this software at a much reduced cost compared to pure commercial development models. Thus if there was a small team which was driving such a project ensuring the quality of the product, then it is possible to get a lot of assistance from the global IT community to make those systems truly exceptional. This is what we see with the Sahana project, which has a core team of 6 people that has built a global community of 60+ participants and contributors. And the FOSS community spirit, philosophy and mechanisms has been a key ingredient in the the successful growth of such a vibrant community for Sahana.
http://newsvac.newsforge.com/newsvac/05/11/05/0553230.shtml http://www.fsf.org/news/social-benefit-award.html
Description ownership areas they have in the relief effort. Database of all requests for support from various locations (camps, hospitals, etc), as well as offers of support from relief providers. Registers all temporary camps, hospitals and locations setup to house the victims of the disaster Database of missing, displaced, dead, orphans etc. (including pictures, finger prints, DNA samples) with advanced search capabilities. Database of all pledges of assistance and attempts to match it to the requests. Also records where the assistance was provided. A collection of key contacts for critical areas during a disaster.
The main systems that were utilized were the Organization registry and the People registry whilst the others were used to a lesser extent. The deployment model included distributed data entry of Gramasevaka data and police data and there were authorized volunteers given access to upload the data into the system coordinated by the universities. Sahana was also loaded on laptops to capture data in the field, however these systems were not deployed due to certain constraints. Overall the deployment model is depicted in the diagram below:
The technology stack used to build Sahana was based on the well known Open Source solution stack, LAMP. This included Debian GNU/Linux for the OS, Apache for the web server, MySQL for the database, and PHP/Java for the Web Application, which makes it pretty much free and open source end-to-end. As the hardware resource requirements of this solution stack is low it was initial deployed on a
standard desktop machine, but subsequently migrated to it own dedicated IBM We also tested and found that it could be deployed on a resource limited PDA such as the iPAQ with OPIE ( a GNU/Linux distro supporting iPAQs) within 64 Mb of RAM. References:
http://www.who.int/entity/hac/events/tsunamiconf/presentations/2_17_private _sector_woodworth_doc.pdf
officially by a government entity in Sri Lanka called the CNO. Sahana is being customized for the Pakistan Quake by NADRA an authorized government entity the government.
2. Build open systems where all relief groups can work together
Any disaster management portal that prevents access to NGOs or civil society or government is a partial solution and yet another parallel silo of data. The parallel silos of data can tend to cause more confusion as no proper data consolidation can take place. In Sri Lanka a move was to host the NGO coordinating authority within the Government's C&C operation, however in Pakistan the government entity responsible for deploying the software seemed reluctant to share it with NGOs (due to their existing policies and security controls).
3. Immediately organize dedicated teams to provide localization, customization and helpdesk services
Thought the Sri Lankan Tsunami requirement was served by a bunch of volunteers working bazaar style, this probably is not the best model to use when customizing and supporting Sahana for deployment in the disaster situation. Disaster environments have requirements and deadlines that are needed ASAP. Every disaster yields different requirements and no matter how well you build your system, there will always be urgent requirements on localization, customization and helpdesk application support. Thus when deploying Sahana it is better to get a dedicated professional team to support it. It is also worthwhile setting up a helpdesk/call center to help users get accustomed to the application and enter data on their behalf (especially in nations with a low IT literacy). In Sri Lanka this was not planned for and though the goodwill volunteer effort generated the application they could not always be relied up to provide 24x7 support.
The Free Software Foundation (FSF) known as one of the two leading organizations responsible for the FOSS movement has created a new award based on humanity, which was directly inspired by Sahana ( Reference: http://www.tectonic.co.za/view.php?id=686 http://www.fsf.org/news/social-benefit-award.html )
Dr Sanjiva Weerawarna received a Redhat User Summit award for inspiring the Sahana system ( Reference: http://www.redhat.com/magazine/008jun05/features/awards/ ) Chamindra de Silva was invited and sponsored to talk on the ICT4Peace panel at UN World Summit on Information Society (WSIS) to talk on the application of the use of ICT during the Tsunami. ( Reference: http://mail.fsfeurope.org/pipermail/wsis-euc/2005November/000635.html ) IBM Crisis team proactively promotes the use of Sahana for disaster management, promoting it both for application after the Katrina Hurricane and the recent Asian Earthquake. ( Reference: http://www.earthquakepakistan.com/images/IBM_CRT_Pakistan_Mission_Re port.pdf ) Sahana was present at the International FOSSSL 05 developer conference ( Reference: http://www.fosssl.org ) Sahana was presented in the 2nd keynote speech in the Emergency Communication Asia 2005 Conference in Shanghai, China. UNDP IOSN network has allocated a section on their portal for HumanitarianFOSS which features Sahana. This section is managed by the Sahana PM/SA. ( Reference: http://www.iosn.net/foss/humanitarian ) Sahana was presented at the IOSN FOSSAP II conference in Cambodia to present Sahana ( Reference: http://www.iosn.net/events/fossap-2005/presentations/casestudies-ict4d/sahana.pdf )
Generalize the application to handle any disaster scenario from a tsunami to and earthquake to a pandemic Adhering to strict LAMP (Linux-Apache-MySQL-PHP) for scalability and ease of deployment. Internationalization of the entire system to allow of L10N to a particular country Build a base application framework to minimize the time to write a new disaster management sub-application Incorporate a good granularity of security features to protect the privacy of data Support additional base architectures such as other databases and OSs Adhere to existing emergency management and humanitarian data exchange standards
To support the above, a disaster management application framework was developed in PHP from scratch that provided the additional features below:
Allows for the independent development of modules/sub-applications by 3rd party groups easy whilst making integration simple
Allows for dynamic plug and play installation and customization of modules required for the disaster with minimal configuration Includes a web setup that automatically launches and set's up the database and the configuration file the first time you run Sahana A HTML template libraries to make writing HTML elements for the forms, menus, etc quick, consistent and elegant A theme and layout engine to allow for the easy modification of the application's user interfaces to cater to the disaster needs A core and flexible data schema shared between application modules
GNU/Linux LiveCD built on Taprobane (Debian based) Dependencies: Apache 1 or 2, MySQL 4.2, PHP 4, Java 1.4 OS: Tested on GNU/Linux Hardware: Tested on Intel PIII upwards License: GPL Target Audience: Government disaster coordination
The Pakistan Sahana Phase I LiveCD can be downloaded from the following location: http://sourceforge.net/project/showfiles.php?group_id=127855&package_id=169330 &release_id=371270
Available Packages
tar.gz source package Debain deb and Redhat rpm binary packages GNU/Linux LiveCD built on Taprobane (Debian based)
Dependencies: Apache 1 or 2, MySQL 4.2, PHP 4 OS: Tested on GNU/Linux, Mac OSX, FreeBSD, Windows XP Hardware: Tested on Intel PIII upwards, Mac Powerbook G4 License: LGPL Target Audience: Government/NGO disaster coordination
Demo: A demo of this release can be found at http://cvs.opensource.lk/sahana2 Download: The PHP application package can be downloaded from http://sourceforge.net/project/showfiles.php?group_id=127855 LiveCD: A LiveCD of the Phase II release has also been created and can be downloaded from the following location http://sourceforge.net/project/showfiles.php?group_id=127855&package_id=169330 &release_id=375026
further refine and pre-deploy Sahana installations globally in order to improve adoption and disaster preparedness when the next large scale disaster strikes. To allow for flexibility we would like Sahana to be able to deployable in a distributed peer-oriented model as depicted below which allows for sensitive data to be protected, whilst allowing data sharing. Additionally we would like it to be accessible in the field and in camps, for which we plan to test hardware units such as laptops with camera and fingerprint scan and PDA with photo and GPS location facilities. Also to be able to work in a disaster where the majority of telecoms infrastructure is destroyed we will be testing access using satellite phones. The new deployment model is depicted below.
The following additional modules will also be built into the system which is in response to the direction provided by the humanitarian-ICT community and problems faced in the Tsunami and Asian quake for which we did not have solutions for: Component Camp Management System (vs Camp Registry) Description A system deployed at the camp to help manage the camp and easily track the victims within it. This will be deployed as a unit with camera and fingerprint scanner Records all damages to property and provides an estimate of the cost of the recovery effort To record burial site information for the benefit of future reference Quickly transforms an synchronizes data from disparate sources into the central database
Description Send mobile alerts (e.g. SMS/MMS) and critical information to those in the disaster area Children tracking, protection and aid management system built specifically for NGO, Terre des Hommes Especially to supplement the Camp management System to track all resource available, their usage and predict when supplies will run out. Generates custom reports and statistics as required for the different organization groups All of access and reporting of events and incidents in the field with GPS and camera enabled PDAs
Other References:
[1] The Open Source Initiative OSI, (2005), (opensource.org), Available: http://www.opensource.org/docs/definition.php [2] LAMP: The Open web platform, (2001) (onlamp.com), Available: http://www.onlamp.com/pub/a/onlamp/2001/01/25/lamp.html [3] Sahana Project, (2005), (sourceforge.net) Available: http://sahana.sourceforge.net/ [4] Document A01 - About SourceForge.net?, (2005), (sourceforge.net), Available http://sourceforge.net/docman/display_doc.php?docid=6025&group_id=1 [5] freshmean.net : About, (2005), (freshmeat.net), Available: http://freshmeat.net/about/ [6] News - Opie - Open Palmtop Integrated Environment, (2005), (opie.handhelds.org), Available http://opie.handhelds.org/ [8] Mozilla Developer Central, (2005), (mozilla.org) http://www.mozilla.org/developer/ [9] CNO Main Page, (2005), (cnosrilanka.org) Available: http://www.cnosrilanka.org [10] Free/Libre Open Source Software, (2005), (wikipedia.org) Available: http://en.wikipedia.org/wiki/Free/Libre_Open_Source_Software [11] The Cathedral & the Bazaar (paperback) By Eric S. Raymond ISBN: 0-596-00108-8