.png?width=2000&name=kognic%20logo-black%20(1).png)
Our team participated in VECS 2023 – the Vehicle-Electronics-Connected-Services conference and expo in Gothenburg. As the autonomy industry evolves, the conversation has shifted from raw data volume to smart data curation. The critical question is no longer "How much data do we have?" but "How efficiently can we turn human judgment into high-quality, validated training data?"
Kognic's CEO and Co-Founder, Daniel Langkilde, delivered the keynote concluding the first day's conference theme, Technologies Enabling Mobility Transformation. His presentation, An Iterative Way of Working Towards Automated Driving, addressed the fundamental challenges facing both Tier 1 suppliers and OEMs:
- How can teams succeed in automated driving when AI capabilities continue to evolve rapidly?
- Why does the traditional approach of specifying product features upfront no longer work when machine learning models continuously improve?
- What makes iterative workflows, strategic partnerships, and continuous experimentation essential for success?
- Why should maximizing engineering velocity be the ultimate goal for any OEM developing autonomous systems
Daniel opened with a direct challenge to the industry:
He emphasized a critical reality in autonomous driving development:
The dataset itself remains the most dynamic variable – a constantly changing picture of the world before a vehicle. The marginal cost of data annotation has traditionally been a major barrier to scaling autonomous systems. To achieve real impact, teams must adopt an iterative process centered on cost-efficient human feedback. This is where Kognic delivers value: we are the price leader in autonomy data annotation, helping customers get the most annotated data for their budget through our unique combination of Platform, Process, and People.
The key insight? Machines learn faster with human feedback. As models become more sophisticated, the bottleneck shifts from raw compute to high-quality human judgment. At Kognic, we enable teams to integrate scalable, cost-efficient human feedback directly into their data pipelines – accelerating the path from raw sensor data to safe, validated autonomous systems.
