Quality assurance · Production

Apollo Tyres reduces manufacturing RCA time by 88% using agentic AI Manufacturing Reasoner on Amazon Bedrock

The problem

Plant engineers at Apollo Tyres manually analyzed curing press bottlenecks using an industrial IoT dashboard, requiring multi-department SME collaboration; subelemental-level root cause analysis was not possible with traditional tools, and delayed insights prevented timely corrective actions.

Workflow diagram · grounded in source
1
Natural language query via UI
trigger
“Users ask their questions in natural language through the UI, which is a Chainlit application hosted on Amazon Elastic Compute Cloud (Amazon EC2)”
2
Primary agent routes query
routing
“The question asked is picked up by the primary AI agent, which classifies the complexity of the question and decides which agent to be called for the multistep reasoning with help of different AWS services”
3
Knowledge base context extraction
ai_action
“Amazon Bedrock Agents uses Amazon Bedrock Knowledge Bases and the vector database capabilities of Amazon OpenSearch Service to extract relevant context for the request”
4
RCA agent multi-step reasoning
ai_action
“RCA agent – This agent for Amazon Bedrock constructs a multistep, multi–large language model (LLM) workflow to perform detailed automated RCA, which is particularly useful for complex diagnostic scenarios”
5
Concurrent explanation and visualization
ai_action
“The primary agent calls the explainer agent and visualization agent concurrently using multiple threads”
6
Results rendered to user
output
“The UI renders the result to the user by dynamically displaying the statistical plots and formatting the records in a table”
7
Real-time anomaly alerting
feedback_loop
“This virtual reasoner also offers real-time triggers to highlight continuous anomalous shifts in DCT for mistake-proofing or error prevention in line with the Poka-yoke approach”
Reported outcome

Manufacturing Reasoner achieved an approximate 88% reduction in manual RCA effort, cutting DCT root cause analysis from up to 7 hours per issue to less than 10 minutes, with targeted annualized savings of approximately INR 15 million in the PCR division alone.

Reported metrics
RCA effort reduction88%
DCT RCA time per issuefrom up to 7 hours per issue to less than 10 minutes per issue
Targeted annual savings in PCR divisionapproximately 15 million Indian rupees (INR) per year
System response time after optimizationapproximately 30–40 seconds
Reported stack
Amazon BedrockAmazon Bedrock AgentsAmazon Bedrock Knowledge BasesAmazon RedshiftClaude HaikuClaude Sonnet
Source
https://aws.amazon.com/blogs/machine-learning/how-apollo-tyres-is-unlocking-machine-insights-using-agentic-ai-powered-manufacturing-reasoner?tag=soumet-20
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Manufacturing Reasoner achieved an approximate 88% reduction in manual RCA effort, cutting DCT root cause analysis from up to 7 hours per issue to less than 10 minutes, with targeted annualized savings of approximatel…

What tools did this team use?

Amazon Bedrock, Amazon Bedrock Agents, Amazon Bedrock Knowledge Bases, Amazon Redshift, Claude Haiku, Claude Sonnet.

What results were reported?

RCA effort reduction: 88%; DCT RCA time per issue: from up to 7 hours per issue to less than 10 minutes per issue; Targeted annual savings in PCR division: approximately 15 million Indian rupees (INR) per year; System response time after optimization: approximately 30–40 seconds (source-reported, not independently verified).

How is this quality assurance AI workflow structured?

Natural language query via UI → Primary agent routes query → Knowledge base context extraction → RCA agent multi-step reasoning → Concurrent explanation and visualization → Results rendered to user → Real-time anomaly alerting.

Related quality assurance cases