Cloud-enabled Video Analytics for Smart Surveillance.

Business Needs

• Software Modernization: Upgrade legacy video gateway software to seamlessly integrate with cloud services, enabling scalable and centralized device management.
• Enhanced Real-Time Monitoring: Implement advanced video analytics to improve asset security, optimize store traffic, and streamline warehouse operations.
• AI-Driven Intelligent Insights: Leverage cutting-edge AI models for object detection, pattern recognition, and incident analysis at both the edge and cloud, facilitating faster and more informed decision-making.
• Robust Data Security: Implement end-to-end encryption and other data protection measures to ensure compliance with privacy regulations and safeguard sensitive information across edge and cloud environments.

The Solution

Designed and developed an AWS cloud-based solution combining both Edge and Cloud AI, with secure data transmission using a multi-layer approach:

• Edge Layer (On-Premises): Redesigned the NVR/Gateway software leveraging AWS IoT Greengrass to enable local execution of AI models for real-time processing. Key features include object detection, facial recognition, and activity recognition, dynamically optimized to maximize in-camera AI capabilities. The solution significantly reduced bandwidth consumption by transmitting only AI-generated insights to the cloud. This design ensures compliance with industry security standards while delivering robust and efficient performance.
• Ingestion and Processing Layer: Integrated AWS Kinesis Video Streams to process on-demand video streams from edge devices and AWS IoT Core for secure, reliable connectivity. Enhanced the processing pipeline by leveraging AWS Rekognition for AI-driven image and video analysis. Utilized Lambda functions for centralized processing, enabling deeper insights through the correlation of Edge AI and server-side AI outputs.
• Storage Layer: Used Amazon S3 for scalable storage of video archives and Amazon DynamoDB for fast, scalable storage of metadata and configuration details related to video streams.
• Delivery Layer: Integrated Amazon CloudFront to provide secure, low-latency access to live and archived streams. Developed a React.js web interface to facilitate interactive viewing by fetching streams directly from CloudFront.
• Security Layer: Leveraged AWS Cognito to manage user authentication, ensuring secure access for both Edge-to-Cloud connections and live video streaming.

Engagement

• Time & Material Contract: Multi-year offshore Time & Material contract with flexibility in team composition changes, resource re-allocation, and release timelines.
• Agile Development: An offshore team of 8 developers and 2 testers adopted Agile/Scrum methodologies, delivering incremental features in 2-week sprints.
• Requirements Management: A dedicated engineer collaborated closely with the customer to align technical deliverables with business objectives.

Project Outcome

• Market Differentiation: The cloud-connected solution provided the customer with a distinct competitive edge, enabling them to offer advanced analytics services to their end users, driving customer satisfaction, and unlocking new revenue streams.
• Enhanced Security: Real-time AI-powered video analytics allowed end users to proactively mitigate security threats, safeguard assets, minimize theft, and create safer environments.
• Operational Efficiency: Retailers and logistics providers leveraged actionable insights to refine store layouts, optimize warehouse traffic flow, and enhance overall supply chain efficiency.
• Timely Product Releases: A highly skilled, co-located engineering team with expertise in diverse technologies ensured efficient development cycles and on-time delivery, facilitating consistent market releases.