Minas Gerais creates pioneering solution for natural disaster prevention with SoftwareOne intelligence and AWS cloud resources

An integrated data platform aids crisis management

Since 2020, Minas Gerais has experienced intensified rainfall-related problems, exacerbated by the climate crisis and global warming. Between 2021 and 2022, heavy rainfall affected the state, prompting a state of emergency or calamity in half of its 853 municipalities and resulting in 70,000 people becoming homeless or displaced.

Faced with this dire situation, the state acknowledged its unpreparedness in addressing these issues. Moreover, there was a lack of preventive measures for the negative effects of the rains, which predominantly impacted vulnerable families living in landslide-prone regions, reliant on social assistance following natural disasters.

In this scenario, Elder Gabrich, Special Advisor at the State Secretariat for Social Development of Minas Gerais, noted that the social assistance policy was unable to respond promptly to those in need, often lacking the necessary resources and financial means. “We had a very atypical 2021 that ended up demanding new actions. That's when we started discussing what could be done to prevent these families from being affected by climatic events in such a severe way,” Gabrich says.

Gabrich emphasizes that during this period, Wesley Matheus, then Special Advisor at the State Secretariat for Social Development of Minas Gerais and now Chief of Staff of the National Secretariat for Monitoring and Evaluation of Public Policies and Economic Affairs (SMA/MPO), initiated discussions on the necessity of preventive measures. The aim was to enable municipalities to identify areas at higher risk of rain impact and the families residing there before the rainy season. “In these families, there are children, elderly, people with disabilities, and people with reduced mobility, and we needed to create an action plan to be able to give a faster response to all these people,” says Gabrich.

Matheus, who already had professional expertise in disaster management, explains that Minas Gerais has a great variability of situations to model a system like this, serving as a reference to other Brazilian states. “In addition to dealing with the rains, the state also has a history of dealing with drought and a component of disaster occurrence. Realising this, we saw that working in the right way would enable us to improve our capacity for both prevention and mitigation and disaster response,” he comments.

Thus, SEDESE had the idea of building a system that would generate information about the risk areas that exist in the state, integrating them with information on where vulnerable families are located in order to prevent severe climatic events. As it did not have the necessary IT structure to develop the solution autonomously, the Secretariat, headed by Matheus' performance as head of the data office, sought the partnership of AWS and SoftwareOne to move the project forward.

“We were able to bring together the state's top management around the theme, and AWS was fundamental for the advancement of the process, as well as the technical support from SoftwareOne, which made it possible, in fact, to structure and develop the system within the AWS cloud in an agile and secure way,” says Matheus.

Innovation on AWS brings hope for the future

To identify disaster-prone areas and communities in Minas Gerais, particularly those at risk of flooding, a new platform was developed using various AWS technologies. Robust security was a key focus, achieved through the implementation of multiple AWS security features.

In developing the solution, SoftwareOne and AWS ensured compliance with Brazil’s LGPD (General Data Protection Law), which mandates the protection of personal data and privacy. The SEDESE system, managing sensitive data of vulnerable families, adhered to these regulations by implementing stringent security measures to prevent unauthorized access and data breaches.

The solution utilized AWS services to create geofences and perform geospatial queries, filtering and analyzing data based on geographic areas and communities of interest. This approach ensured compliance with LGPD’s digital sovereignty requirements, safeguarding citizens’ data privacy and security.

The technologies utilized to accomplish this include:

• AWS Lambda - To integrate external sources (APIs) to receive public alerts from the National Institute of Meteorology (INMET), which indicate weather forecasts such as rainfall intensity, low or high humidity, among others, and to integrate the CadÚnico database – a record of low-income families in Brazil.
• AWS S3 - To store the data using the Lakehouse medallion architecture and the encryption features to increase security.
• AWS Key Management Service - To handle encryption keys securely.
• AWS Network ACLs - To control inbound and outbound traffic at the subnet level.
• Security Groups - To restrict traffic between the Internet and the Amazon VPC.
• AWS IAM - To manage users, including setting permissions and roles.
• AWS Glue (a serverless data integration service) - To collate and integrate data from multiple sources.
• AWS Athena - To execute query data.
• AWS Location Services - To locate data in the platforms.
• AWS Amazon Managed Workflows for Apache Airflow - To orchestrate actions and reduce the complexity of maintaining and monitoring the solution.
• AWS QuickSight dashboards - To present the data to public bodies, defense agencies, and social welfare departments.
• AWS SNS - To send messages to those impacted or in emergency situations.

Leveraging PaaS (Platform as a Service) resources, the final solution was designed to enhance cost control, simplify development and maintenance, and ensure scalability, resilience, and availability.

The result was the creation of a solution that allows SEDESE to georeference people and control data through dashboards, with the map pointing out when there are meteorological alerts. In practice, the system can search municipalities to determine the number and location of people at risk based on weather forecasts, facilitating proactive planning by civil defense and social assistance teams, as well as the distribution of humanitarian aid.

“Based on the proposal designed by SEDESE, SoftwareOne was the first company in Brazil to develop this type of solution, which is already common in first-world countries, especially in the USA. In this sense, the solution will allow the government entities of Minas Gerais to work with proactive and preventive plans, which was not possible before,” comments Leal.

The Mapeia Minas project has already been delivered by SoftwareOne to SEDESE and the solution is in use. The initiative will also be expanded to all other states that are part of COSUD (South and Southeast Integration Consortium), which will be able to use the same tool.

“Although we are working with data samples at this first moment, the project has met our expectations. We have received strong support from the SoftwareOne team to overcome our challenges and continue with the development of the system,” says Gabrich.

He further adds that the first major milestone aimed for is to be able to make a periodic shipment of information to the municipalities of Minas Gerais in the next rainy season, which occurs from October to March. “We want to offer information about risk areas, which families live in that region, and the profile of these families so that they know the problem and, with that, know what the level of difficulty is to solve it in a flood situation, for example,” emphasises Gabrich.

Leal reinforces that this movement will be very important for the country by allowing the population and government entities to take precautions with more time. “Thus, it is possible to reduce the population's losses, directly impacting the quality of life of society,” Leal concludes.