How Tata Elxsi detects industrial safety risks in seconds on AWS
Tata Elxsi built IRIS, a serverless real‑time safety platform that detects hazards in seconds by running edge‑AI on cameras and streaming only key frames to AWS.
Why Now
The story follows Tata Elxsi’s release of IRIS, a new industrial safety platform built on AWS, announced on 22 Sep 2026.
What Happened
IRIS uses AWS IoT Greengrass on GPU edge servers (e.g., NVIDIA Jetson AGX Orin) to filter frames and run a lightweight model before sending only safety‑relevant frames to Amazon S3. The filtered frames are streamed via Amazon Kinesis Data Streams to downstream inference, correlation, and alerting services. In production, edge filtering cuts cloud‑sent frame volume by 70–80 %.
Why It Matters
By processing video locally and sending only essential data, IRIS reduces bandwidth and latency, enabling near‑real‑time hazard alerts that scale with camera count. This can lower human monitoring costs and improve compliance in large industrial sites.
The Limitation
The article focuses on a single customer deployment; performance may vary with different camera setups, network conditions, or safety‑critical environments.
What You Can Do
Try deploying a small edge‑AI demo with AWS IoT Greengrass and Kinesis to see how much data you can filter out before sending to the cloud.
Source
Read original sourceWhy we picked this
Core AI content: real-time industrial safety platform using SageMaker AI and edge computing.