Case Study

Validation of analytics models which detect the warranty claims for locomotive engines

The client is a Fortune 500 locomotive company that designs and manufactures engines especially in the Heavy Equipment sector.

Key Challenges

The client pays its customers’ warranty claims for the locomotive engines. Detecting these faults took three to five months and another three months for resolving

The process led to a lot of manpower and monetary impact with huge amount factored for this task on an annual basis

To tackle this huge revenue blocking, the client needed a way to predict future failures and warranty claims

The Impact

30^%

reduction in budget provisioned for warranty claims by early prediction of the faults after the go-live of Machine Learning (ML) algorithm models in two of the engines.

The Solution

Infosys Advanced Analytics Program: Validation of analytics models

26% of major defects were detected due to well-defined strategy. This ensured that the program’s aim of ensuring around 85% accuracy for prediction was achieved

Looking for a breakthrough solution?

TALK TO OUR EXPERTS

26% of major defects were detected due to well-defined strategy. This ensured that the program’s aim of ensuring around 85% accuracy for prediction was achieved

Raw data from the sensors placed in the locomotive engines were fed into the ML algorithm/model to detect faults and issues and predict the warranty claims in advance. Infosys Advanced Analytics Program validated this model and made sure the predictions were accurate.

From a ‘no validation program’, the QA strategy has helped to bring a robust validation system to the client’s data analytics landscape.

Infosys implemented a well-defined QA strategy to authenticate the analytics models
Team built customized automation solution to speed up execution in one-week sprint duration
QA validation strategy helped in identifying core defects. ~26% of major defects were detected due to well-defined strategy. This ensured that the program’s aim of ensuring around 85% accuracy for prediction was achieved

Strong knowledge of niche technologies

Infosys’ strong knowledge of niche technologies like ML Algorithm-Weibull, Occurrence Monitoring, SubPopulation, and the QA skills helped in end-to-end exhaustive validation.

WHITEPAPER

Enabling QA through Anaplan model testing

This white paper examines the key best-practices that will help organizations benefit from seamless planning through successful Anaplan testing.

Experience

Insight

Innovate

Accelerate

Assure

Application Development and Maintenance

Business Process Management

Consulting Services

Incubating Emerging Offerings