How Amazon builds and deploys autonomous robots
summary
Amazon Robotics describes building an industrial data fabric on AWS (Redshift, S3) feeding an AI layer (Amazon Bedrock, Quick Suite) and an Agent Core to automate manufacturing decisions. Specific deployments: Irma cuts inventory investigation from 3–4 hours to 36 minutes and boosts inventory accuracy ~15%; Artemis gives 80% first-pass repair resolution and halves mean time to repair for complex robots. Computer vision flags ~34% of eligible defects; Siemens Tecnomatics digital twins and Nvidia simulation enable 320% throughput growth with 60% faster response times.
AI-written summary · about
chapters (7)
- 0:00Introduction and background
- 3:19Digital transformation strategy
- 4:46Key AI-driven projects
- 10:04Test execution and Artemis
- 11:28Digital twin for optimization
- 14:57Training and scalability impact
- 16:24Business outcomes and lessons
related talks
26mInnovate with AWS: Enabling industrial ops at Amazon fulfillment centers
26mInnovate with AWS: Enabling industrial ops at Amazon fulfillment centers
28mBuild smart robots and operate them easily with AWS Robotics
30mBuilding, testing and operating robots with AWS
15mAI at Amazon
29mHow Amazon builds AI-enabled perception for robots
30mCustomer centricity in manufacturing with operational improvements
33mReducing the environmental impact of agriculture with robotics