What can we do now with Digital twins, Simulation and AI for Disaster Management? Prevention, Preparedness, Response and Recovery

Disasters, whether from natural hazards or man-made, cost lives, jobs, and have huge financial and economic costs for organizations, individuals, and societies. The human costs are staggering too. Over the past two decades, more than 7,000 major recorded disaster events have claimed 1.23 million lives and affected 4.2 billion people (often more than once), reports the United Nations Office for Disaster Risk Reduction. The total damages from disasters caused by natural hazards globally have risen by 800 per cent to $167 billion a year in the last decade from $18 billion a year in the 1980s, according to the World Bank.?

Countries, regions, municipalities, businesses and other organizations have made progress in managing disaster risks – whether floods, hurricanes, wildfires, earthquakes, tornados, ice storms or terrorist attacks – by pivoting from responding only when a disaster strikes to a more proactive and comprehensive approach. This means addressing each of the four phases of disaster planning and management:?

• Prevention to minimize the impacts of future disasters
• Preparedness to prepare for managing the crisis
• Response to save lives and minimize immediate impact
• Recovery to restore activities and services

Together, #digitaltwins , #simulation and #artificialintelligence (#AI) can help governments and organizations plan and more effectively manage each phase to prevent or minimize human and financial damage that result from any of these events.

Simulate and remodel a city or region to reduce vulnerability to future disasters – from digital twins to meta twins

Prevention is about being proactive: identifying potential hazards and devising mitigation strategies to minimize the damage and loss of human life that would result from a disaster. Digital twins offer accurate and incredibly detailed #3D replicas of a neighborhood, city, region or country. In Presagis ’ new VELOCITY 5D platform, we use AI algorithms not only to speed up and perfect digital twin creation, but also to allow the digital twin to simulate, predict and show the effects of different types of disasters on an environment. The outcome is that organizations can clearly visualize causes and consequences to better anticipate, interpret and understand how a disaster would affect their city or region.?

For example, AI makes it is possible to identify which buildings would be destroyed given the characteristics of a certain hurricane and predict the damage. It can also simulate different flooding scenarios to help identify danger zones and specific vulnerabilities, and to identify evacuation and relief routes. Understanding and predicting is a very important first step, but AI can do more. It can artificially alter an existing digital twin to find the best way to minimize the impact. We can name such altered twins “meta twins” instead of digital twins, as they are representative, plausible and accurate versions of a digital twin – they just don’t yet exist. AI algorithms can mimic and replicate at scale the complex and statistical nature of our reality at a level of precision that would not be possible by any human or expert.

Using digital twins and AI-based simulation models (vs. real-life tests) is both very safe and inexpensive, unlocking endless possibilities. AI algorithms can then be used in processing and analyzing the data needed to evaluate each of the scenarios in terms of benefits and costs, and guide decision-making and choices with respect to investing and allocating resources, whether they be constructing levees and seawalls, emergency shelters, reinforcing buildings or any other mitigation measure.

Preparing and training emergency personnel for what to do in a future disaster

Preparedness is a process in which governments, businesses and other organizations plan and train for what they will do in the event of disaster to ensure the highest level of readiness. In the preparedness phase, it’s critical for organizations to train crisis leaders and emergency-response personnel for decision-making, logistics, and quick, effective, and efficient action in different disaster situations.?

Digital twins, simulation and AI can be used for training emergency personnel – emergency managers, firefighters, police, paramedics, emergency medical responders and other medical professionals – by simulating various disaster scenarios in their city or region with high fidelity. By running through different virtual scenarios, emergency personnel have an opportunity to practice and learn safely and inexpensively how to react appropriately and effectively as individuals and as a team to crises they likely never have experienced. With AI, it’s also possible to run live training simulations that include both real people and autonomous agents (simulated people), that behave as real people likely would. Emergency personnel could also do offline training, with autonomous agents, on their own schedule.

Through AI and digital twins, one can develop optimal strategies and principles, as well as an action plan with personnel training to manage and respond to a crisis. The disaster can easily be tailored to the scope, size and type of one’s choosing and different scenarios can simulate its impacts on the specific local environment. AI can also be useful in determining the best rules to apply for logistics and deployment of emergency personnel, equipment and other available resources. All this information and intelligence can help guide evidence-based, financially responsible decisions about optimizing or increasing the budget for emergency personnel, supplies and equipment to be better prepared for future disasters.

Using AI, simulation and digital twins to improve disaster response?

During a disaster, time is of the essence. Emergency leaders and first responders need to have good situational awareness of what’s happening and assess the immediate consequences to make the best decisions in responding to a disaster.

A digital twin with simulation is like having a live, changing Google map, which shows what is happening in the affected area in real time. In the case of flooding, for example, this virtual replica can show the direction of the water and water levels moving through a city and its neighborhoods. With AI, satellite images, video and other data can be used to quickly identify damaged buildings and provide a complete picture and damage assessment. Emergency response managers and personnel can also use these tools to help save lives, and treat or prevent injuries, by identifying and alerting people at risk, and mapping the best routes for rescue and evacuation, whether for dispatching ambulance or helicopter rescues.

In disaster situations, it can be difficult for emergency responders to accurately assess their own level of risk as they assist during the event. AI and digital twins can help emergency managers accurately evaluate the risk level for emergency workers, guiding them in carrying out their work safely and effectively and preserving the emergency services’ capacity to rescue citizens.

In certain types of disaster situations, like the massive earthquake and tsunami that damaged and caused a loss of power at the Fukushima Daiichi Nuclear Power Plant in Japan in 2011, it was too dangerous for emergency personnel to access the disaster site. This catastrophic event illustrates how AI, remote control technologies and autonomous systems (including robots) can be used to prevent the spread of damage and mitigate the disaster, and to protect workers and the public by reducing the risk of radiation exposure. Digital twins can dynamically connect and communicate the required information to autonomous systems on the terrain to guide such systems.

Digital twins, simulation and AI can also help disaster response by analyzing various sources of data to predict the path and severity of extreme weather events. For example, a 2021 report in the journal Nature Communications, showed how an AI-enabled tsunami forecasting method used real-time observation data – including seismic earthquake data – and lessons learned from past tsunami experiences to support quick and appropriate evacuation. Another example from Google Deepmind, published in 2021, showed how AI can quickly and accurately “nowcast” (i.e. in the next 1-2 hours) rain locally. AI can also be used for real-time evaluation of fire propagation risk for wildlands and rural areas.?

Dynamic blueprints for restoring services and activities in a transformed environment?

In the post-disaster recovery phase, governments and organizations must set priorities for restoring essential services like food, clean water, power, transportation, and healthcare first, and less essential services and activities subsequently. The full recovery process can take weeks, months, years or even longer in the case of a catastrophic event, such as Hurricane Katrina, which struck the Gulf Coast and devastated New Orleans in 2005.

To restore services and activities effectively, efficiently, and safely, recovery leaders should have an accurate and detailed understanding of what’s changed in the environment. Digital twins can be used to reconstruct an updated, detailed replica of the altered environment including damage to infrastructure, buildings, terrain and the population. The goal may not be to restore the pre-disaster environment but to remodel, rebuild and restore the altered environment in an optimal, timely and cost-effective way. Together, digital twins, simulation and AI can be used to simulate different reconstruction and restoration scenarios and inform decisions on implementing the best solutions before engaging in costly new development and rebuilding. Presenting a sound and detailed, evidence-based recovery plan to granting agencies – bolstered by utilizing digital twins and AI – may increase the chances of success in applying for recovery funding.

The post-disaster recovery phase is also an opportunity to learn lessons from the disaster, and for governments and organizations to think about what they need to do differently to minimize damage and loss of life next time. Digital twins and AI have an incredible capacity to track and analyze what happened during the crisis and why. This intelligence and information can be used during recovery to help improve disaster prevention and preparedness – as part of a continuous #disastermanagement cycle -- by making buildings more resilient and updating training for emergency services personnel.

Together, #digitaltwins , #simulation and #AI are powerful tools to help governments and organizations plan and prepare for major disasters, and to see their effects before they happen. This trio of next-generation technology can also help guide #decisionmaking and the implementation of effective strategies for responding and recovering from disasters to minimize damage and protect people’s lives. #Naturaldisasters are, unfortunately, increasing in frequency and severity around the world.

At Presagis with VELOCITY 5D, we are leveraging the latest advances in these technologies to help governments, businesses and other organizations increase their disaster readiness . I am excited and proud of what we are accomplishing at Presagis expanding the limits of what these technologies can do and bringing to market solutions that reduce the risks and dangers to the many communities and people affected by disasters.

#disastermanagement #climatechange #deeplearning #AI #simulation #metatwins #generativeai #reinforcementlearning #syntheticdata