The AI×Drone Wide-Area Infrastructure Security Optimization Survey Project has been selected for a subsidy from the Ministry of Economy, Trade and Industry.
- Midori Suzuki
- Aug 22
- 4 min read
Federative Republic of Brazil / Crime Prediction AI x Drone Wide-area Infrastructure Security Optimization Research Project
Singular Perturbations Inc. (Headquarters: Chiyoda-ku, Tokyo; President and CEO: Kajita Mami; hereinafter, SP) announced on August 22, 2025, that its project to optimize wide-area infrastructure security in the Federative Republic of Brazil using crime prediction AI and drones has been selected for the Ministry of Economy, Trade and Industry's FY2024 supplementary budget for the Global South Future-Oriented Co-Creation Project Subsidy (Small-Scale Demonstration/FS Project).
This project will demonstrate a new wide-area security model that combines our proprietary crime prediction AI "CRIME NABI" with an autonomous drone patrol service, targeting large privately owned areas such as mines and plants in the Federative Republic of Brazil. In areas where human patrols are difficult, drones will automatically patrol along optimal routes based on AI predictions, aiming to dramatically improve security efficiency and strengthen security standards.
------------------------------------------------------------------
* FY2024 Supplementary Fund for Global South Future-Oriented Co-Creation Projects
(Small-scale demonstration/FS project) Special website
------------------------------------------------------------------
Issues and needs of wide-area infrastructure security
In emerging countries such as Brazil, there is a high demand for security at industrial infrastructure sites covering vast tracts, such as mines and oilfield plants, to prevent the devastating damage caused by resource theft and destruction of equipment. However, this demand cannot be fully covered by human patrols. For example, the mining sites of CSN Mining, a major Brazilian mining company, and the oil well facilities of Petrobras, an oil company, cover areas ranging from tens to hundreds of square kilometers, making it an urgent issue to prevent theft of resources such as copper wire and fuel, as well as illegal entry. In fact, when the Brazilian police used SP's crime prediction system, CRIME NABI, in their patrol operations, it was reported that the theft of copper wire, a type of infrastructure material, decreased by 69%, demonstrating the high demand for enhanced security at large privately owned sites.
・Examples of potential customers: Mining development companies, oil and gas plant operators, businesses that own large-scale agricultural plantations and forest conservation land, companies that manage ports and logistics centers, etc.
An innovative security model combining crime prediction AI and drone control
In this demonstration, SP will link data from its proprietary crime prediction AI, "CRIME NABI," with the drone traffic management system (UTM) to build a new security model in which autonomous drones automatically select and fly optimal patrol routes based on risk maps predicted by the AI. CRIME NABI predicts future crime risks based on past crime patterns and environmental data, and proposes routes that prioritize patrols of high-risk locations. This enables drones to automatically patrol vast areas based on AI predictions, potentially dramatically improving patrol efficiency, security coverage, and response speed in the event of an emergency. SP established a local subsidiary in Brazil in 2023 and has a track record of verifying the effectiveness of CRIME NABI with multiple police agencies. The company will apply this knowledge to this security model for private infrastructure.
Potential benefits:
- Improved patrol efficiency: Drones automatically patrol high-priority areas based on crime predictions, enabling efficient security even when there is a shortage of manpower.
- Expanded security range: By utilizing fixed-wing drones, etc., it is possible to monitor a wide area at once, covering the entire vast site area that was previously not possible.
- Enhanced response speed: When a risk is detected, a drone flying nearby will immediately check the scene, enabling a faster initial response than before.
Proven proprietary technology and a track record of international expansion
In a patrol demonstration by the state and municipal police in Belo Horizonte, Minas Gerais, Brazil, the number of thefts of traffic lights and copper cable from communication cables was reduced by approximately 68.5% after the introduction of CRIME NABI. This is a notable example of the significant contribution of the company's AI-based crime prediction technology to crime prevention in the field, making CRIME NABI the world's only AI security route optimization technology proven to be effective in real-world situations. CRIME NABI is an AI solution based on a crime and accident prediction algorithm developed by SP. It analyzes past crime and environmental data to accurately predict when and where crimes and accidents are likely to occur, visualizing risks in advance and ensuring maximum safety with minimal resources. This unique technology has been refined through national research projects, including those with NICT and NEDO, and its high effectiveness has been confirmed, including the deployment of services to improve the efficiency of patrol and security camera monitoring operations for state and municipal police in Brazil, where crime rates are high. To date, CRIME NABI is used by 14 police agencies in Latin America, including Brazil, demonstrating the global acceptance of the company's technology.
Vision: Taking on the challenge of transforming global security infrastructure
This demonstration will serve as a starting point for the commercialization of drone security services for wide-area infrastructure in Brazil and Latin America, with the aim of establishing a global standard for security infrastructure in the future. During the demonstration, the company plans to collaborate with local companies and security agencies to verify effectiveness and operational flow, leading to commercialization of the service from 2026 onward. Centered around the integration of Crime Nabi (crime prediction AI) and UTM (drone control), the company will expand its SaaS + operational service model and broaden its application to B2B outdoor security, including mining, energy, ports, and logistics. SP will leverage its strengths in AI and on-site implementation capabilities cultivated through research and development to establish a next-generation security infrastructure that integrates "prediction → patrol → detection → response," and will continue to contribute to social safety together with its collaborative partners.
[Company Profile]
■Singular Perturbations Inc.
Head Office: 5F KS Floor, Resona Kudan Building, 1-5-6 Kudanminami, Chiyoda-ku, Tokyo 102-0074
Representative: Representative Director and CEO Kajita Mami
Established: August 3, 2017
Business: Development and provision of crime and accident prediction AI "CRIME NABI"
Awards and adoption history: Winner of B Dash Camp 2024 Sapporo / Selected as one of Forbes JAPAN's "7 startups to watch in 2025" / Adopted for public projects such as NEDO, NICT, IDB Lab, and JICA
■ SINGULAR PERTURBATIONS, SERVICOS DE TECNOLOGIA LTDA
Address: Av. Paulista, 807 – Cj. 1922 – Jardim Paulista – São Paulo – SP – CEP 01310-100
Representative: Representative Director and CEO Takamasa Kurachi
Established: October 31, 2023
Business: Development of security solutions utilizing crime prediction AI "Crime Navi"
■ Contact Information
Singular Perturbations Inc. Public Relations Officer
