Key results
The company
Aumovio
aumovio.comHardware and software solutions for software-defined and autonomous vehicles.
Result highlights
- 40% operational cost reduction
- 10% improvement in model accuracy
The story
An automotive technology provider navigating the industry's shift from discrete control units to centralized high-performance computing architectures.
Software testing relied on physical hardware availability, causing development cycles to drag on for weeks. Additionally, engineers struggled to efficiently identify the critical 1% of relevant data hidden within massive multi-modal datasets needed for training autonomous driving models.
The organization built a cloud-based platform that allows developers to provision virtualized electronic control units (vECUs), enabling software testing without physical hardware. The system leverages AWS Graviton processors for simulation and generative AI to automatically isolate relevant data for model training. This architecture decouples development streams, allowing teams to iterate on vehicle software as rapidly as smartphone applications.
Quotes
“The solutions we have developed together with AWS allow us to reduce our operational costs by around 40% and improve the accuracy of algorithym by 10%.”