Cyclone Idai Reflection: Data Science In Disaster Preparation And Relief

Hi! My name is Njabulo Sandawana and am the founder of Ant Analytica. Ant Analytica is a Big Data Analytics startup based in Harare. At Ant Analytica we leverage data science and Analytics to help companies build automated solutions and provide insights that lead to sustainable growth. The recent disaster that occured in the Eastern Highlands as a result of cyclone IDAI got us thinking of ways we could levage Data Science for better disaster management and preparation.

We believe that technology and Big Data Analytics aided by Artificial Intelligence can transform disaster relief efforts by enhancing prediction and preparation abilities and by accelerating response time and enhancing responders’ ability to operate efficiently even when resources are scarce.

Here is a write up on this important subject:

The rapidly increasing frequency of disasters has become a menace to human habitation across the globe. Effective disaster risk reduction and management can be achieved through the deployment of geospatial data for all the phases of disaster management, including prevention, mitigation, preparedness, vulnerability reduction, response and relief.The lack of information can delay response times which could potentially bias the distribution of Aid during disasters. Hence information (data) becomes invaluable in providing the best and most efficient means of disaster management and relief. Beyond that, data analytics can help predict disasters and model appropriate measures to deal with the disasters accordingly.

Crisis Mapping

In order to assess the flood disaster, in extent and coverage and ascertain the level of damage in the environment and the number of communities affected by the flood, Rapid Mapping of the affected area should be carried out . This sort of mapping is effective for providing information for rehabilitation; mitigate and prepare against future occurrence.In order to mount effective responses, emergency managers need accurate maps that show the extent of damage, predictions for its potential spread, and detailed data on the movement of people and resources. With the use of smartphones we are can use geospatial data that goes down to the level of individuals, as well as maps showing key infrastructure and up-to-date damage assessments created on the fly, in order to manage response efforts.

Forecasting and predicting

Data analytics allows for the identification of important population subsets such as the elderly, physically disabled/ handicapped in communities and make the necessary adjustments to the type of aid provided in rescue missions and prevent unnecessary loss of time and more importantly life associated with a one-fits-all approach. Satellite remote sensing can be used for timely and near-real-time disaster detection. Making use of multitemporal remotely sensed imagery, which is captured over the same location at several points in time. The major goal of this early monitoring phase is to define a boundary that delimits the affected area, so that the preliminary info can be generated.

Damage Assessment

Making use of remote sensing imagery at low resolutions and large scales provide quick initial assessment of the impacted areas. A more detailed damage assessment of buildings and traffic networks would require higher resolution and 3D data to provide information about the degree and intensity of damage. A task easily carried out through the use of UAVs,which can off the higher resolution needed for detailed analysis by surveying and 3D mapping the area toassesthe damage.We strongly believe that they are new opportunities provided by emerging technologies and they need to be understood for the betterment of disaster relief management. For example, with the technological advancements, it is understood that big data tools can process large amounts of disaster-related data(traditional humanitarian data as well as user generated data) to provide insights into the rapidly-changing situation and help drive an effective disaster response. Big data analysis can help test particular frameworks of resilient supply chains in the context of disaster relief activities.

Data Driven Disaster Response

Response Every disaster provides an enormous amount of data . Mining information from previous disasters, of icials and responders can collect insights that help forecast future incidents. A combination of sensor data collection, surveillance and satellite imagery, big data analytics will allow mission-critical areas to be surveyed and assessed. Knowing, for example, that a particular area has been flooded, and by how much, provides highly useful data for mapping out flood-prone zones and planning for where to store key rescue resources nearby. Through AI flood patterns can be predicted and help to bring greater precision and accuracy to response ef orts. We can use Drones to collect data for contextual mapping when it comes to tackling wildfires. And satellite imagery be useful in understanding level of destruction and check conditions of roads to understand obstacles to access.

Summary

In an era of data-producing connected devices, what matters is not only the quantity of data collected but also how these data are managed and analyzed. Technology and big data analytics aided by artificial intelligence are transforming disaster relief efforts by enhancing prediction and preparation abilities, and by accelerating response time and enhancing responders’ ability to operate efficiently even when resources are scarce. Equipped with the right data management strategy, governments can be smarter about preparing for disasters and improving aid efforts—when every second counts.