Scala at Audi Optimization of assembly lines
summary
Pipeline that maps configurator features to binary vectors and clusters them to find feature-vector clustering and part-to-feature mappings. Implements an in-memory variant store and distributed compute on a Leimen cluster to enumerate and filter millions of valid car configurations, then counts required parts and identifies a minimal set of assembly variants. Shows actual deployment into Audi infrastructure, scales to production data and produces concrete assembly sequences that reduce redundant part handling.
AI-written summary · about
chapters (8)
- 0:00Scala at Audi intro
- 2:16Assembly line challenges
- 8:05Historical context
- 13:27Production use cases
- 19:07In-memory performance
- 24:02Clustering feature mapping
- 27:10Instrumentation implementation
- 30:49Results and outcomes
related talks
28mAI-driven query plan optimization in big data processing tools
01h 03mScaling up machine learning without tears (and what do programming languages ha...
50mFull-Stack AI in Java: From Cassandra to CarPlay - A Production SaaS Story
19mDrive transformation through machine learning
24mScalable GenAI in Production
49mBuilding Machine Learning Pipeline Using Aerosolve
34mComplex Machine Learning Pipelines Made Easy
11mAWS On Air ft. BMW & BCG