Humanitarian-FOSS Case Study On Disaster Management

Humanitarian-FOSS: Case Study On Disaster Management ISCRAM 2007 Harbin, China, Aug 26-27 2007 Chamindra de Silva Sahana, Project MC Chair and Director Acting Exec Director, Lanka Software Foundation IEEE-CS SL, Founder Member Apache, Committer Agenda • • • • • • • • Humanitarian Response Multi-Agency Coordination How can I.T. help Disaster Coordination? Free and Open Source Software (FOSS) Open Standards Disaster Management: Sahana Project How it works: the software, the community Lessons Learned Scope • Humanitarian Response • In the “large scale” – Is the practice of saving lives and alleviating suffering in natural and man-made disasters – Local emergency management teams have insufficient capacity or are part of the disaster themselves – You need multi-organizational, multi-national groups to respond to the disaster effectively – Lot of dynamic information exchange requirements Humanitarian Response “Large Scale” Disasters Government Relief Government Ministries, Police, Army, Fire fighters, + Well trained + Authorized coord + Accountable +/- Big picture relief (e.g. national security) - Procedures create bottlenecks - Overloaded UN, NGOs Red Cross, OCHA, CARE, WHO, Sarvodaya, etc + focused on people + trusted to accept aid +/- less well trained + accountable - fragmented / disconnected - sometimes competing Local Relief Spontaneous volunteers, corporate village communities, friends and family + first responders + lot of capacity + instant aid - not trained - focus unknown - not accountable Effective Collaboration and Coordination! Relief Victims Example: Asian Tsunami • Coordination Issues – Over a million displaced people to help with food, shelter, clothing – 100,000s missing people to track – 100s of NGOs and Relief groups to coordinate – Multitudes of aid and • Situation supplies to track and – 15 Countries Affected handle – Managing all relief workers – 230,000 Deaths and volunteers and – 1.7 Million Displaced applying their skills effectively... Overwhelming amount of information to be collected, shared and distributed to the right people at the right time How Can IT Help? • Scalable management of information – databases, search, reports and dissemination • Empowerment of responders with Information – The right information at the right time – help the trusted work efficiently • Mediation and coordination – Help responders for multiple agencies self-distribute themselves and provide a balanced response • Dynamic and Live reports for decision makers – Situation awareness What is FOSS? • Free and Open Source Software – e.g. Linux, Firefox, Apache, Red Flag Linux -> • Open Source – Is a set of principles and practices that promotes access to the design and production of goods and knowledge (software => Access Source Code) • Free as in Speech, not Free as in Beer – Freedom to access, run, modify and redistribute the software as you wish => A global public good • A Global Community – Anyone can participate • A melting pot of diverse skills and vocations – A meritocracy on contribution • “Action is the most valued thing” Why is FOSS relevant? • Freely Available to deploy – No discrimination on access (Red Cross CC #2) – Ability to “leave technology behind” and scale it up • Rapid customization to actual needs with code – L10N and integration (RC CC #7) – Building local capacities & self-reliance (RC CC #6) – Build on each others work by including it in project • Open system => Transparent and trustworthy – Better acceptance than “foreign” proprietary systems • Countries/NGOs can collaborate to build this solution – Get the best minds from the world to participate in building the software • This should be a global public good Called Humanitarian-FOSS or H-FOSS Open Standards • “Open Standard” – Royalty Free – Open Process for developing and maintaining standard – Available Open Source implementation e.g. Internet Protocols – From TCP/IP • Why is it relevant to Humanitarian Software – Enables system to system integration – Royalties will constrain use => Reduce the effectiveness of holistic disaster response • Emergency Response Open Standard Groups – OASIS EM TC (EDXL, CAPS) – W3C EM Ontology and Interop XG The Sahana Project Example • What is it? – A portable web based tool with sub-applications designed to address the common disaster coordination issues – Free and Open Source • Main Goals – Bring Efficiencies to Disaster Coordination with IT solutions – Facilitate the effective information exchange between a diverse groups of people (NGOs, Government, EM groups and victims) – Primary focus is to help victims Sahana Deployment • CNO, for Tsunami in Sri Lanka – 2005 – Officially deployed and track 26,000 families • NADRA, for Asian Quake in Pakistan – 2005 – Officially deployed and integrated to NADRA (Pakistan Government) to track all victims • 3 Disaster in Philippines – 2006 – Officially deployment to track all victims with by Philippines Government + pre-deployment • Yogjarkata Earthquake, Indonesia – 2006 – Deployed by ACS, Indonesian Reliefsource • For local government in a key state in US • And now: Peru Earthquake Coordination Challenges Finding missing, injured and displaced people “Please, please help me find my missing parents / child / relative” Missing Person Registry • Centralized Bulletin Board of Missing People • Tracks person descriptions, pics, seeking details, seeker information • RSS feed of missing people Coordination Challenges Who is doing what, where? “We are an NGO with medical services, where can we help?” “In which region is there a lack of support and relief services?” Organization Registry • Relief Contact Directory • Matrix reports by sector, affected regions • Drill down location hierarchies to identify gaps and owner • Enables “self-organization” Coordination Challenges Matching pledges of aid to requests from the field “We need x number of tents not clothes!” “We have medical supply x. Who needs it the most?” Request/Pledge Management • A tracking system from requests for aid from the field and pledges of aid by donars • Helps connect donors to requesters independantly => “self-distribution” Coordination Challenges Building the Situation Map “What is happening out there? What areas are affected?” “I would like to report X happened at location Y” Situation Map / Camp Registry • A collaboratively built map of the disaster situation • Ability to mark camps, victims, incidents, warnings, etc with pictures and notes • Track facilities and numbers in shelters Other Modules • Other Modules / Features – – – – – – – Volunteer Management Inventory Management Evacuation Management Modules Alerting / SMS Module Localization Many more built by the community You can build a module for submission • Sahana is a flexible modular framework – You can choose which modules you which – And you can then customize as much as you wish Sample Deployment Model • Centralized for authorized coordinators – Inter-connected through the Internet • Data collection distributed – Paper Forms, other manual means • Data entry distributed or centralized – Can be done with volunteers • Accessible interface Key Features – Focus on simplicity and intuitiveness – Translated to your local language • Low Hardware costs – Can be run on most old computer – Software is free • Mobility (Internet not needed) – Can be access on a PDA browser – Can run from a USB disk (program + data) – Can run on a laptop without Internet access • Open Standards Supported – CAP (PFIF, EDXL Pending) Sahana Portable App Mobile operation through USB disks Works without internet access Data exchange to peers and central server through USB flash disks Who Builds the Tool? A Global Community (200+) A diverse group of Emergency Management Practitioners, Humanitarian Consultants, Academics and Software Developers − Key Organizational Contributions from LSF, IBM, SIDA, Google, ISCRAM, Universities: Uo Colombo (Sri Lanka), Uo Moratuwa (Sri Lanka), Uo Maryland (US), Trinity College (US), Tilburg University (Netherlands) Structure: − Board − PMC − FOSS Committers − How is the project run? • Direction and Requirements – Live disasters, NGO and Gov stakeholders – Community of Humanitarian Consultants, Emergency Management Practitioners, Academics – Sahana Board (Sahana Strategy) • Research – Faculty and students from Sri Lanka, US, Netherlands • Development & Testing – Project Management Committee ( Project Governance ) – Committers (Development & QA) • Deployment – IBM Crisis team, LSF, Bluepoint (Philipines), NADRA (Pakistan) Lessons Learned • Technology is not the end goal – Helping victims and responders is the end goal – So don't let you judgment get clouded with your passion for technology • Mis-understanding that Open Source means data is not protected – – Code is free, but data is protected Real question: • how much can you trust the data with a product? • Build Simple Resilient Response Solutions • Ensure it is accessible to your target group – Reduce number of points of failure and have many backup strategies – Encourage simplicity to reduce learning curve Summary • Problem of “large scale” disaster coordination • Humanitarian-FOSS. • Open Standards • Sahana Disaster Management System • Application in Philippines • You can participate! Welcome to Participate! • Everyone is welcome to join the community • Contribute Code • Contribute your leadership in the Domain • Contribute to standards – Join a the Humanitarian-ICT community – http://www.yahoogroups.com/humanitarian-ict – Join a H-FOSS project like Sahana – http://www.sahana.lk • Contribute by using and deployingg – Join a standards group like W3C Disaster Ontology group – http://esw.w3.org/topic/DisasterManagement APPENDIX • Get the system officially authorized Few Lessons Learned on Deployment – Promoted through the central relief coordinating authority designated by the Government • Encourage data centralization and sharing – Between government, NGO, UN, etc – promote integration to trusted groups • Securing the Perimeter vs Accessibility – Solid ACLs and perimeter security is needed • Pre-deployment is best – Otherwise L10N and customization on the fly • Build simple, resilient response strategies – 1 Laptop, 1 Wireless LAN, Car battery • Give practitioners the OS they find comfortable – Often a Windows XP + WAMP + Sahana – Or Windows XP + Sahana PortableApp Lessons Learned on using FOSS • Passionate community of contributors – The merger of two passionate/strong willed communities • Easy acceptance from user orgs – Free(dom), Transparent, Transfer of ownership • Scepticism – 24x7 Support, Information Security Humanitarian-FOSS • A Definition – The application of Free and Open Source Software to help solve problems with regard to alleviating suffering and saving lives • Recognition of Concept – Free Software Foundation Award for Social Benefit inspired by Sahana (and won recently) – UNDP IOSN • Community – 150+ Emergency Management Practitioners, Humanitarian Consultants, Academics, Students – Humanitarian User Groups (HUGs) When is Relief Chaotic? • Signs of a Large Scale Disaster – When the relief needs overwhelms existing Gov and Emergency Management resources • Local emergency management infrastructure is can also be destroyed − When multi-Agency response is critically needed   Multitudes of NGOs and foreign aid groups arrive simultaneously to help the relief effort Too many parallel and independent relief tasks to coordinate lack of bandwidth => coordination delays => lack of information => lack of trust − When the official disaster coordinator is overwhelmed   Overwhelming amount of information to be collected, shared and used to empower responders and victims