Enhancing Disaster Response Capacity for PDRF on AWS
Case Study: Philippine Disaster Resilience Foundation (PDRF)
Customer Overview
The Philippine Disaster Resilience Foundation (PDRF) is a nonprofit organization dedicated to strengthening disaster preparedness, response and recovery efforts across the Philippines. PDRF relies on geospatial data and analytics to coordinate multi-stakeholder disaster response operations involving public and private sector partners, and humanitarian organizations.
As part of its long-term digital transformation strategy, PDRF has partnered with Help.NGO to modernize its cloud infrastructure and expand its disaster response capabilities beyond geospatial systems. The engagement includes Amazon Web Services (AWS)-based initiatives focused on resilient communications, edge computing, AI-enabled operational support and secure cloud-native collaboration, thereby creating a scalable technology foundation for future disaster response operations.
Customer Challenge
PDRF's disaster response operations depend on access to geospatial information, timely updates from field personnel, and reliable communications between deployed teams and coordination centers. However, these capabilities can be affected by aging infrastructure, fragmented reporting workflows, and unreliable connectivity during disasters.
The existing HANDA Geographic Information System (GIS) environment runs across on-premise and AWS infrastructure and includes an older Windows-based ArcGIS Server environment. This created several concerns around long-term support, security, maintenance and the ability to upgrade to newer versions of ArcGIS.
Furthermore, information from field teams is often communicated separately from the GIS environment. Responders may need to report their location, safety status, mission progress, or incidents while simultaneously carrying out their primary responsibilities. During a disaster, these updates can come from multiple teams and locations and may not always follow the same format.
This created several operational challenges:
Aging GIS infrastructure that is increasingly difficult to maintain and support
Risk of disruption to HANDA services during upgrades, failures, or disaster operations
Field updates that may be reported in different formats across teams
Additional workload on responders who need to provide updates while carrying out response activities
Difficulty in quickly consolidating responder locations, safety information, incidents, and mission status
Limited ability to connect field information with the broader geospatial picture used by coordination teams
Dependence on public cellular and internet infrastructure that may be degraded or unavailable during disasters
Limited ability to access communication tools, maps, and other operational systems when field locations lose backhaul connectivity
PDRF needed to modernize the systems supporting these workflows without disrupting ongoing disaster response operations.
Goals/Objectives
The objective of the engagement is to improve the reliability and maintainability of PDRF's disaster response systems while making operational information easier to collect and use.
The initiative focuses on three areas:
First, the HANDA GIS environment was modernized through the deployment of a new ArcGIS Server 12.0 environment in AWS. The migration was designed to move PDRF away from its aging Windows-based environment while maintaining access to existing GIS data, maps, applications, and services throughout the transition.
Second, PDRF with Help.NGO introduced NextCloud on Amazon S3 to store their data and have role-based access to their data.
Third, PDRF implemented a Wickr-based check-in chatbot that allows field responders to provide structured operational updates through a communication tool already used by teams. The chatbot can collect information such as responder location, safety status, mission progress, operational readiness and incident reports.
Together, these initiatives are intended to improve how PDRF collects, manages, shares, and accesses operational information before and during disaster response activities.
Customer Recommendation
Help.NGO recommended a phased modernization of PDRF's disaster response environment rather than replacing all existing systems at once.
For HANDA, a new ArcGIS Server 12.0 environment, built in AWS alongside the existing production environment, was proposed. GIS data and services would be migrated and validated in phases, beginning with historical and static information, before moving current and live services. The existing environment would remain available until the new platform has been tested and accepted.
For field operations, Help.NGO proposed an AI-assisted check-in chatbot integrated with Wickr. The chatbot provides responders with a simple way to submit structured operational updates while allowing AWS services to process and route that information for use by coordination teams.
This approach allows PDRF to improve its existing systems while minimizing disruption and changes to the tools and workflows already familiar to its teams.
Solution Optimality
Several approaches were considered for modernizing PDRF's disaster response systems:
Continuing to operate the existing HANDA environment with incremental upgrades, which would retain the maintenance and support limitations of the aging infrastructure
Replacing the existing GIS and communication tools with a new platform, which would require larger changes to established PDRF workflows and increase migration and adoption risk
Modernizing HANDA separately while continuing manual field reporting, which would address the GIS infrastructure but not the challenges around collecting consistent operational information from responders
Modernizing the existing systems on AWS while introducing new capabilities through tools already used by PDRF
The selected approach allowed PDRF to improve existing systems without requiring teams to adopt an entirely new operational platform. HANDA remains the primary geospatial environment, while Wickr continues to provide secure communications for field teams. The NextCloud-based architecu provides a way to keep critica data available to the right people even locally.
This also allowed the capabilities to be introduced independently. The HANDA environment could be migrated in phases, the Wickr chatbot could be validated through a limited Proof of Concept (POC).
Solution Overview
Help.NGO is supporting PDRF with an AWS-based architecture covering both geospatial services and field reporting.
HANDA GIS Modernization
The new HANDA environment includes:
A supported SQL Server environment and new ArcGIS enterprise geodatabase
AWS networking and infrastructure deployed through CloudFormation
ArcGIS Pro for preparing and republishing existing GIS services
Phased migration of historical, static, current, and live GIS data
Incremental updates to ArcGIS Online maps and applications
Parallel operation of the old and new environments during migration
Backup and rollback procedures before retirement of the legacy environment
Wickr Check-In Chatbot
The field check-in solution uses:
AWS Wickr for secure communication with responders
Amazon Bedrock to support the AI-assisted chatbot
Amazon EC2 to host the WickrIO and application bridge
AWS Lambda for check-in and reporting automations
AWS Systems Manager Parameter Store for configuration and version management
Structured workflows for location, safety, mission progress, readiness, and incident reporting
The chatbot is designed to reduce the amount of manual reporting required from field personnel while producing more consistent operational information for coordination teams.
This system is designed to provide PDRF teams with local connectivity and access to critical applications in the event of connection loss. When connectivity is restored, field data can be synchronized with the centralized AWS environment.
Users send check-in messages through AWS Wickr to a WickrIO and FastAPI bot on an EC2 host inside a VPC, which pulls its configuration and secrets from SSM Parameter Store and passes each message to an Amazon Bedrock Agent. The agent calls the CheckInLogic Lambda action group, which logs check-ins to DynamoDB and emails recipients via Amazon SES, while Amazon Managed Grafana turns the DynamoDB data into a dashboard that users can open.
Users reach Nextcloud on an EC2 instance in the VPC’s public subnets through an Internet Gateway, with its metadata kept in a PostgreSQL RDS database in the private subnets and its S3 access key held in AWS Secrets Manager. Nextcloud stores user files in an S3 bucket as its object storage, and VPC endpoints let the private subnets reach S3 and AWS Systems Manager, used to manage the instance over HTTPS, without that traffic going over the public internet.
Implementation
The HANDA migration is being carried out gradually to reduce operational risk. The existing production environment remains available while the new AWS environment is deployed, configured, and tested.
Historical and static GIS data will be migrated first. Once these services have been validated, current and live services will be moved in coordination with PDRF. ArcGIS Online references will be updated individually rather than through a single cutover.
The Wickr chatbot focuses on structured check-ins, incident reporting, location reporting, wellness and safety updates, and operation under realistic connectivity conditions.
Keeping the work phased allows PDRF to test each capability before it becomes part of normal disaster response operations.
Results and Impact
The combined work gives PDRF a stronger foundation for managing information during disaster response.
Modernizing HANDA reduces PDRF's dependence on aging infrastructure and provides a supported environment for its geospatial data and applications. The phased migration approach also reduces the risk of losing access to important GIS services during the transition.
The Wickr chatbot addresses a different part of the same operational problem by making it easier for responders to provide consistent updates from the field. Coordination teams can receive structured information about responder locations, safety, incidents, and mission progress without requiring field personnel to use a separate reporting system regardless of internet connection availability.
As these capabilities mature, the structured information collected from responders can also support PDRF's broader geospatial, situational awareness, and disaster resiliency workflows. This creates the potential for coordination teams to see both the geographic conditions surrounding an emergency and the status of teams operating within those areas.
The result is a more complete operational picture for PDRF during disaster response, built around systems that can be maintained, expanded, and integrated over time.
Challenges and Opportunities
The project involves modernizing systems that support active disaster response operations, which means changes need to be introduced without affecting PDRF's ability to access GIS services or communicate with teams in the field.
For the HANDA migration, one of the main challenges is moving from the existing ArcGIS environment to ArcGIS Server 12.0 while preserving existing GIS data, services, ArcGIS Online maps, applications, and integrations. A direct cutover would create unnecessary operational risk, particularly if migration activities were to overlap with an active disaster response. To address this, the old and new environments will operate in parallel while data and services are migrated and validated in phases.
Field operations present additional challenges. Responders may be working in areas where public cellular networks or internet backhaul are degraded or unavailable. At the same time, teams need to communicate, access maps and operational information, and provide updates on their location, safety, mission progress, and incidents. Requiring separate tools or manual reporting processes can unnecessarily add to the workload of personnel already operating in difficult conditions.
These challenges also create opportunities for PDRF to improve how information is managed across disaster response operations. Modernizing HANDA provides a supported foundation for PDRF's geospatial services, the Wickr chatbot provides a more structured way to collect information directly from field teams.
As these capabilities develop, PDRF can explore ways to connect structured field reports with its existing geospatial workflows. For example, responder locations, incident reports, and mission updates could eventually complement the maps and operational information already available through HANDA, thereby giving coordination teams a more complete view of activities in affected areas.
Security and Compliance Considerations
Compliance Identification
The solution requires secure handling of geospatial and operational information used during disaster response activities. This includes controlling access to HANDA GIS services and protecting information submitted by field responders through the Wickr check-in workflow, as well as data stored or processed at field locations and synchronized between the edge environment and centralized AWS services.
Because the systems may contain responder locations, safety information, incident reports, and other operational data, access and data handling need to be managed according to PDRF's security and privacy requirements.
Compliance Implementation
Help.NGO implemented the following controls:
Controlled access to ArcGIS services and supporting AWS infrastructure
Encryption of data at rest and in transit
Role-based access and least-privilege permissions for users and administrators
Wickr end-to-end encryption for responder communications
Data minimization for information collected through chatbot workflows
AWS Identity and Access Management (IAM) controls for access to supporting cloud resources
Version-controlled configuration for the chatbot and Amazon Bedrock Agent
Logging and monitoring of supporting AWS services
Backup, testing, and rollback procedures for the HANDA migration
Structured change and deployment procedures before production services are transitioned
AWS KMS for encryption key management
Secure synchronization between field and centralized AWS environments
Security Assessment and Gap Analysis
Initial State
HANDA operates on aging Windows-based infrastructure that requires modernization
GIS services need to remain available while the environment is upgraded
Field check-ins and operational updates may be communicated through separate or manually structured workflows
Field teams may operate in areas with intermittent connectivity
Operational information from responders and geospatial systems is not yet part of a single integrated workflow
Identified Gaps
Long-term security and support risks associated with the existing HANDA environment
Risk of operational disruption if GIS services are migrated without adequate testing and rollback procedures
Inconsistent structure in field check-ins and operational reporting
Need to minimize unnecessary sensitive information in responder reports
Need for secure and controlled processing of chatbot messages and supporting AWS services
Limited integration between structured field information and PDRF's broader geospatial workflows
Loss of connectivity risks loss of access to all pertinent data and critical applications
Implemented Improvements
Deployment of a new supported ArcGIS Server 12.0 environment in AWS
New supported SQL Server environment and ArcGIS enterprise geodatabase
Phased migration and validation of GIS data and services
Parallel operation of the existing and replacement HANDA environments during transition
Secure Wickr-based workflows for responder check-ins and operational reporting
Data minimization and privacy guardrails for chatbot interactions
Controlled access to AWS services based on least-privilege principles
Logging, monitoring, version management, and operational runbooks
Structured testing and validation before wider production deployment
Alignment with AWS Well-Architected Framework
Security: Controlled access, encryption of data at rest and in transit, least-privilege permissions, and secure Wickr communications help protect geospatial and operational information. Data minimization and access controls are also incorporated into the chatbot workflows.
Reliability: Phased migration, parallel operation, backups, and rollback procedures reduce the risk of disruption to HANDA services.
Performance Efficiency: AWS compute resources support the processing requirements of ArcGIS workloads, while services such as AWS Lambda provide lightweight processing for chatbot automations and operational reporting workflows.
Operational Excellence: Infrastructure deployment through AWS CloudFormation, documented migration procedures, monitoring, version management, and phased validation provide a more consistent approach to managing and updating PDRF's disaster response systems.
Cost Optimization: Moving away from aging infrastructure reduces the need to maintain legacy systems and allows AWS resources to be sized according to workload requirements. The use of managed and serverless services for supporting chatbot functions also limits the need for additional dedicated infrastructure.
Broader Strategic Partnership
The ArcGIS modernization forms part of Help.NGO's broader AWS partnership with PDRF. Building on this work, additional initiatives have expanded the collaboration into resilient communications, edge computing, and AI-enabled operational support. This has created a roadmap for future disaster response capabilities.
Comments