AWS re:Invent 2025 - Real-time insights for smart manufacturing with AWS Serverless (CNS375)

Watch	Topic	Details
YouTube	Introduction to the Problem	The speaker shares a personal anecdote about baking cookies to illustrate the problem of disconnected data and expertise in manufacturing. The problem of unplanned downtime in manufacturing costs the top 500 manufacturers globally 1.4 trillion US dollars annually, equivalent to the GDP of a nation like Spain.
YouTube	Identification of Key Issues	Three main issues are identified: data silos, skill gaps, and the visibility of problems when they occur. Data silos refer to the disconnected data sources from different machines in a production line. Skill gaps refer to the lack of knowledge sharing among experts and junior operators. The visibility of problems refers to the delayed response due to multiple data sources and lack of expertise to analyze them.
YouTube	Proposed Solutions	The solution involves creating a unified data lake using AWS Garnet framework to integrate different data sources. A GenAI application deployed over Amazon Bedrock is proposed to bridge the skill gap and provide real-time data insights. Automation using AI is suggested to take actions based on data insights without requiring senior expertise on the ground.
YouTube	Implementation Details	The implementation involves using AWS IoT Core, Lambda functions, and S3 to consolidate data into a data lake. Actions such as updating production line settings are facilitated through API Gateway. Data preparation and fine-tuning of the model are done using AWS Step Functions and SageMaker.
YouTube	Model Fine-Tuning and Deployment	The model is fine-tuned using structured data and imported into Amazon Bedrock for unified API access. A web application is created to expose a chatting interface with the model, utilizing the Converse API for flexibility and scalability.
YouTube	Continuous Learning and Improvement	The system supports continuous learning through feedback mechanisms, allowing for model updates based on operator feedback. QR codes are provided for further implementation details on Garnet framework, small language model implementation, and machine learning pipeline.

Watch on YouTube

Description

Manufacturing facilities need instant access to operational data for smarter decision-making. This serverless solution leverages AWS Lambda, Amazon Timestream, and Amazon S3 to deliver real-time, multilingual safety protocols and technical insights to the factory floor. This architecture eliminates infrastructure overhead while enabling immediate analysis of production data, helping manufacturers optimize operations and maintain safety compliance through data-driven insights.

Learn more:
AWS re:Invent: https://go.aws/reinforce.
More AWS events: https://go.aws/3kss9CP

Subscribe:
More AWS videos: http://bit.ly/2O3zS75
More AWS events videos: http://bit.ly/316g9t4

ABOUT AWS:
Amazon Web Services (AWS) hosts events, both online and in-person, bringing the cloud computing community together to connect, collaborate, and learn from AWS experts.
AWS is the world's most comprehensive and broadly adopted cloud platform, offering over 200 fully featured services from data centers globally. Millions of customers—including the fastest-growing startups, largest enterprises, and leading government agencies—are using AWS to lower costs, become more agile, and innovate faster.

#AWSreInvent #AWSreInvent2025 #AWS