View All Events
Virtual
Americas
Workshops
Physical AI
Autonomous Vehicles
Debugging Physical AI Models at Scale with Multimodal Data - August 11, 2026
This event has ended, but you can still catch up! Watch the on-demand recordings and register for our future events.
Aug 11, 2026
9:00 AM - 10:00 AM PST
Online. Register for the Zoom!
Host: Prerna Dhareshwar
Machine Learning Engineer / Product & Customer Success, Voxel51
About this event
As robotics and autonomous vehicle teams move from traditional perception models to end-to-end Physical AI systems, understanding model behavior is becoming harder than ever. These models ingest synchronized inputs from cameras, sensors, and other data streams, but their decisions can be difficult to explain, reproduce, and improve.
Watch this on-demand workshop on how multimodal data workflows in FiftyOne help teams visualize, search, and curate complex Physical AI datasets at scale. We show how teams can work natively with raw MCAP recordings across synchronized video and sensor streams, tag and index segments of interest, and query for similar scenarios across their datasets using embeddings-based search, faster than playback-only visualization tools allow.
You'll learn how to use multimodal data to investigate questions like: where did a gripper close around an object, when did a pedestrian appear on a crosswalk, or where else in your dataset does this same action occur — and how can you find every similar moment across your dataset?
Designed for robotics, AV, and machine learning teams, this session will show how FiftyOne helps turn multimodal data into a scalable workflow for model evaluation, debugging, and improvement.
Host
What you’ll learn:
- How multimodal data workflows in FiftyOne help inspect, search, and debug complex Physical AI datasets
- How to work with synchronized video and sensor data to understand model behavior
- How to tag segments of interest manually or automatically via the SDK, and how derived events (like a gripper state change) get generated from your sensor signals
- How to query for similar failure scenarios across large datasets
- How to uncover patterns behind model failures faster than playback-only visualization tools
- How to turn multimodal data into a scalable workflow for model evaluation, debugging, and improvement
Who this is for:
- Robotics and autonomous vehicle teams building or deploying end-to-end Physical AI systems
- Machine learning engineers working with synchronized camera, LiDAR, radar, or sensor data
- Teams responsible for debugging, evaluating, or improving Physical AI model behavior
- Data scientists and ML engineers who need to find and analyze rare failure modes at scale
Resources
© 2026 Voxel51 All Rights Reserved