Case Study: Reaktor achieves nearly 10× faster model development for an autonomous ferry with Valohai

Urban waterways The next generation of autonomous transportation

Reaktor, the Finnish software powerhouse, set out to build a prototype autonomous ferry to tap underused urban waterways and relieve city transport strain. Their challenge was training a self-steering deep learning model quickly and reliably under a tight schedule with limited on‑prem hardware — a problem they addressed using the Valohai platform.

Valohai provided a scalable platform and cloud GPU orchestration that cut system and infrastructure setup time by 80% and increased model development speed almost tenfold, enabling overnight training instead of a week. Those gains let Reaktor move from limited on‑prem resources to cloud GPU clusters, finish the prototype ahead of schedule, and continue enhancing the system (for example by adding LIDAR) to advance autonomous waterways.

Reaktor