Supplier Recovery Claim Automation at Cummins Inc

Supplier Recovery Claim Automation Veronica Heber, Senior Reliability Analyst Neha Kichambare, Reliability Analytics Specialist Cummins Inc.

Agenda Introduction and Background Challenges of Previously Existing Process Automating the Report Using JMP Working With Data Interacting With Users Results Demonstration 2

Introduction Background Mapping warranty claims to key suppliers: Partnerships with suppliers Standardize warranty sharing agreements Leads to improving quality for the customer Cummins Dependability Need for Automation: Improve cycle time Standardized and Transparent reporting Implement best practices

Process of Mapping Claim to Supplier From warranty claims with replacement part information, use the defined algorithm to identify the supplier

Process of Mapping Claim to Supplier The algorithm integrates information from multiple databases to automate mapping of claims to suppliers Input Automated mapping process Output Claims supplier 1 Claims supplier 2 Claims supplier 3 Warranty Claims Mapping algorithm Replacement part number original part number (supersession list) OEPL1 (Number of claims and $CoC6 assigned to supplier) Part number + engine plant + build date supplier name GPS2 Supplier names unique supplier name SIM3 Service kits component part number CPIF4 Engine-failcode causal part noun name Dictionary5 1 Original Equipment Parts List (OEPL) 2 Global Purchasing System (GPS) 5 Dictionary: Engine-failcode to most likely causal part noun name data table developed using CMI expert inputs 3 Supplier Information Management (SIM) 4 Central Product Information File (CPIF) 6 Cost of Coverage (CoC)

Challenges of Previous Process With the previously existing tools and resources, there are four key challenges

Solutions Through this Project To overcome these challenges, the team developed an algorithm and implemented it in JMP 1 2 3 4 </> </> Obtain missing information by leveraging multiple data sources through Query Builder Automation in JMP can handle the high volume of data Link disparate databases with common identifiers Translate non-standard nomenclature through Text Explorer

Query Builder Create and run customized queries against the various systems Join tables on date ranges Eg, date in service is within the time the part was sourced causal_records_firstmatch = New SQL Query( Connection( "JMP" ), QueryName( "causal_records_firstmatch" ), CUSTOM SQL( "SELECT * FROM causal_nonkit_claims t1 LEFT OUTER JOIN GPS_ALL_GOLDENSUMMARY t2 ON ( ( t1.REPL_PART_7 = t2.GPS_PARTNUM_7 ) AND ( t1.PLANT_CODE = t2.SOURCE_CODE_PLANT ) AND ( t1.FAIL_BUILD_DATE >= t2.EFFECTIVE_DATE ) AND ( t1.FAIL_BUILD_DATE <= t2.NEW_EXPIRATION_DATE ) )" ) ) << Run;

Text Explorer Variations in supplier names occur due to free-text entry fields Text Explorer helps identify name re-coding and supplier recovery opportunities

Graph Builder Use Graph Builder to create custom graphs Save as PowerPoint to share with users and suppliers

Notifications to Users Remind users to periodically update certain lookup tables Helps with maintenance, keeps the report current Results and Information from the report Provides immediate feedback for the user about the report status

Add-In Allows for efficient report distribution and maintenance Easy for users to run Automatically check for updates

User Interface Allows user input Provides information about the report

Results JMP s capabilities for working with data: Save countless hours of manual processing Help standardize warranty agreements for key suppliers In 2017: 48 suppliers Through 2019: 200+ suppliers Improvements in efficiency, scalability, and best-practices JMP s tools for interacting with users: Enable us to deliver exactly what our users need with custom interface and user input Promote continuous improvement and best-practices through notifications to users and free-text analysis

Demonstration

Questions? 16