Transforming public safety in Thailand: Deploying AI police cyborgs with AWS Cloud solutions

This post was written by Amazon Web Services (AWS) and Security Pitch, an AWS Partner.

Law enforcement in Thailand has faced significant limited resources during the country’s festival periods, for example during the Songkran festival. During the Maha Songkran World Water Festival in 2025, more than 550,000 people attended the celebrations organized at Sanam Luang in Bangkok, while Silom and Khao San roads drew approximately 170,000 and 197,000 people, respectively. As public gatherings grow in size, the Royal Thai Police (RTP) must extend their reach and optimize their resource allocation to ensure safety across high-density event areas. To address this challenge, an innovation solution was deployed: an AI-powered police robot, the AI Police Cyborg 1.0.

Developed by Security Pitch Co., Ltd, an AWS Partner, the AI Police Cyborg 1.0 was implemented during the 2025 Songkran festival in the Nakhon Pathom province in Thailand. This advanced robotic officer helped monitor large crowds and enhance public safety, using the power of Amazon Web Services (AWS) Cloud infrastructure and AI.

Figure 1. The AI Police Cyborg 1.0, developed by Security Pitch and operated by the Royal Thai Police. The name of the robot is “Pol. Col. Nakhonpathom Plod Phai,” which translates to “Nakhon Pathom is safe.”

Key features of the AI police cyborg

The AI Police Cyborg 1.0 possesses cutting-edge features and functionality, including:

• 360-degree surveillance: The robot is equipped with smart AI cameras that monitor the surroundings in real time.
• Facial recognition: The AI system automatically identifies and selects the optimal facial frame from the video feed to match with watchlists.
• Suspect tracking: The robot monitors and tracks suspects across the event based on facial recognition.
• Advanced search capabilities: The robot searches for individuals using facial features, clothing, body type, and gender.
• Weapon detection: The robot identifies dangerous objects like knives and guns while ignoring harmless ones like water guns during the festival.
• Behavior detection: The AI system analyzes video feeds to identify and distinguish between aggressive or suspicious behavior and signs of medical emergencies or distress. It provides real-time notifications to appropriate responders, such as police officers for security threats or medical personnel for health-related incidents.
• Unified surveillance network: Multiple video sources—including AI cyborg cameras, CCTV networks, and drone footage— are integrated into a centralized surveillance ecosystem. It enables cross-platform tracking, centralized monitoring, and real-time coordination through a unified command center dashboard.

Building real-time AI systems for the AI police cyborg

Security Pitch chose AWS to build the AI systems powering the police cyborg, using AWS’s advanced AI services, agility, scalability, security, and resiliency.

Security Pitch needed to build multiple machine learning (ML) models to support each task (such as facial recognition and weapon detection). Amazon SageMaker Model Training provides high-performing ML compute infrastructure and can scale GPU infrastructure up or down. The ML models are deployed using Amazon SageMaker Inference, which can scale the model deployment based on the traffic, and with less operational burden. To effectively track the resources and model behavior violations overtime, Security Pitch use Amazon SageMaker Model Monitor to monitor metrics like data quality, model quality, bias drift, and feature attribution drift. Moreover, Security Pitch uses Amazon SageMaker Pipelines for automating model training and deployment as part of their machine learning operations (MLOps) automation.

To effectively search and compare the faces of suspects and individuals on watchlists from the Royal Thai Police’s crime system, facial vectors are recorded, embedded, and stored in Amazon Aurora PostgreSQL-Compatible Edition for secure storage, search, and comparison. Amazon Simple Storage Service (Amazon S3) is used for scalable storage of video feeds, ML model artefacts, and images.

Security Pitch uses Amazon Simple Queue Service (Amazon SQS) and AWS Lambda for queuing and orchestrating the flow of data between devices and each AWS services, Amazon Simple Notification Service (Amazon SNS) for delivering critical notifications to law enforcement, and Amazon Route 53 for routing and accessing the data internally within domain name system service from the polices’ devices and command centers to ensure that all critical alerts reach officers instantly and securely.

The following diagram shows the solution architecture.

Figure 2. Core AWS architecture behind AI Police Cyborg 1.0

This integrated approach enables the AI police cyborg to process live video feeds from the cyborg, CCTV cameras, and drones; analyze crowd behavior; and identify persons of interest and weapons with high accuracy. The architecture supports automated, serverless workflows that reduce response times and improve situational awareness for law enforcement.

“Public safety is no longer reactive—it must be proactive, predictive, intelligent, and always-on. AWS Cloud gives us the power to deliver AI policing at national scale, without compromising speed, security, and citizen trust,” said Pakorn Thongjeen, CEO and founder of Security Pitch Co. Ltd.

Tangible results of AI police cyborg

The deployment of AI Police Cyborg 1.0 during the Songkran festival achieved very positive results. The Provincial Police Region 7 reported over 20 arrests based on outstanding warrants, including:

• More than five suspects arrested at the Aircraft Market in Mueang Nakhon Pathom for drug offenses, theft, attempted murder, and firearms violations.
• Over three suspects apprehended at Central Salaya shopping mall for drug distribution and illegal gambling operations.

Compared to previous years, the AI-assisted policing led to a 30 percent increase in warrant arrests and a 20 percent reduction in reported incidents during the festival period.

Future roadmap: AI Security Cyborg 2.0

Security Pitch plans to debut RAION V2 Sentinel or AI Security Cyborg 2.0 globally in September 2025 in West Virginia, USA. This next-generation cyborg will feature enhanced AI capabilities and natural language interaction, powered by AWS AI services. New features will include:

• Real-time, two-way communication for citizen assistance
• Multilingual support for tourist information and translations
• Improved emergency reporting and incident filing capabilities

The AI Security Cyborg 2.0 aims to transform the role of robotic law enforcement from surveillance to community engagement and support.

Conclusion

The successful deployment of AI Police Cyborg 1.0 in Thailand demonstrates the power of combining advanced robotics with AWS Cloud and AI services. This innovative solution not only enhanced public safety during a major festival, but also showcased the potential for AI-driven law enforcement tools to optimize resource allocation and improve community interaction.

As we look toward the future with AI Security Cyborg 2.0, the partnership between Security Pitch and AWS continues to push the boundaries of what’s possible in public safety technology, promising even greater advancements in the years to come.