Building Self-Adaptation Capabilities for a Socio-Cyber Physical System of a Disaster-Affected Region
Transporting medical resources during disasters is a particularly wicked problem, since any plan to travel between clinics or triaging station must be adaptable to changes in staffing or logistics. This is especially true in a highly uncertain and volatile environment where road conditions may not be known before transportation begins.
Our research tackles the question
How can disaster managers analyse resource distributions and staffing alternatives in highly dynamic and uncertain disaster scenarios?
To reason mathematically and precisely about disasters, the region is characterised as a socio-cyber-physical system. The system keeps track of resourcing per location within the region. Humans play critical roles in socio-cyber-physical systems both as sophisticated sensors who notify managers of frequent and unexpected changes in resource demand and as actuators who provide system functionality such as transporting the physical resources.
Our research uses a four phase Monitor, Analyse, Plan and Execute (MAPE) control feedback loop to compute and carry out redistribution plans to transport resources where they are needed. Disaster resourcing requirements are constraints developed during the preparedness stage of disaster management to be evaluated against current resourcing by the MAPE loop’s analysis phase. Quantitative verification techniques are used on probabilistic models synthesised from the system to reason with uncertainty. Managers can further evaluate planning alternatives such as trade-offs among staffing assignments before carrying out resource distribution operations.
Our research builds on a successful international collaboration of expertise in public health and software engineering. The use of a self-adaptive resourcing approach is new to the domain of disaster management and presents a significant technical research contribution.
Download Roads tool
Install the Roads Source Code [PDF, 141.4 KB]
Interpretation of ROAD outputs and Scenarios [PDF, 229.1 KB]
- Kenneth Johnson (AUT)
- Sam Madanian (AUT)
- Javier Cámara (University of Málaga, Spain)
- Roopak Sinha (AUT)
- Dave Parry (Murdoch University, Perth, Australia)
- Akshay Raj Gollahalli (AUT)
- Mathew St. Martin (AUT)
DCT Funded Project 2021
For more information contact
Dr Kenneth Johnson
Software Engineering Research Laboratory (SERL)