Using an AI-first approach to bring the next-generation self-driving solution to commercial trucking.
- Founded: 2021 in Toronto, Ontario, Canada
- Size: ~100 employees (Series A)
Drop-in-place tagging system
Foxglove API endpoints
On-demand data streaming
Panels and custom extensions
Rapid development and debugging
Self-driving companies have historically approached autonomy software development with hand-crafted and hand-tuned algorithms, but Waabi is trying something different. They’re taking a much more scalable approach by harnessing AI to power all aspects of their development. They’ve built a groundbreaking AI-based simulation system and virtual driver for autonomous trucking operations – one that can generalize much better across different real-world scenarios.
As Waabi began its self-driving journey, the team recognized the need to navigate dense data, organize test results, and resolve issues quickly. While laying the proper groundwork would accelerate all future development, Waabi knew that building a complete robotics observability platform would be a significant investment of time and resources.
By adopting Foxglove and the MCAP file format, Waabi was able to hit the ground running. The team took advantage of Foxglove's existing foundation to accelerate their development and testing processes and to streamline collaboration.
Understanding their data to get to market faster
By being a “late mover”, Waabi has witnessed firsthand how other robotics teams have operated – and often failed. Historically, robotics companies have each built bespoke versions of similar tools to organize data, triage incidents, and debug issues. But creating a world-class developer tool suite that would meaningfully accelerate work could easily take years – from staffing a full-time internal team to building out a workable MVP and maintaining the software with bug fixes and new features. These autonomous driving companies were burning valuable resources on duplicated work that didn't align with their core competencies or ultimate goal.
As Waabi began developing their virtual driver software, they wanted to avoid making the same mistake. They integrated Foxglove to access tooling that would work out-of-the-box and with minimal setup. Instead of reproducing Foxglove features in-house, Waabi unblocked their engineers right away and let them hit the ground running doing what they do best – applying their cutting-edge AI and ML research to build autonomy and simulation software.
We knew we wanted to leverage off-the-shelf developer tooling where possible – especially because we saw how costly it had been for other companies to reinvent the wheel. By adopting Foxglove, we were able to focus on our unique differentiators.- Daryn Nakhuda, Head of Software at Waabi
Recording drive data in a common format
Waabi adopted MCAP, Foxglove’s open-source file format, as a unified data format to represent their driving data. They knew they wanted an efficient append-only file format to avoid on-vehicle overhead, but also wanted time-indexing and topic partitioning to retrieve arbitrary subsets of the data on demand. Paired with Foxglove, the MCAP format helped Waabi realize these goals without their engineers having to create custom formats and data pipelines.
With MCAP, the team had a common data language that they were speaking. Instead of having to worry about different subsets of their multimodal data being recorded in different formats, they could store these files in the same place, organize them in a consistent way, and inspect them with the same suite of tools. Instead of wasting time on converting data into different formats or switching between tools for various tasks, everyone on the team could simply focus on uploading data, debugging issues, and improving their autonomy software.
Streamlining cross-team collaboration
While Waabi first started using Foxglove to streamline development for its autonomy engineers, numerous teams across the company now use the platform for a wide variety of tasks – including inspecting driving logs, identifying key scenes for triage, and visualizing data while the system is on the road.
Waabi engineers have also been annotating their data with Foxglove events to identify interesting scenarios robots encounter on the road. By evaluating metadata across multiple events, Waabi can easily categorize them into relevant buckets for further analysis. This tagging feature has helped the team augment their simulation scenarios test suite and improve their autonomy models.
Composing rich debugging and visualization workspaces in Foxglove, and sharing them with team members, has also accelerated Waabi’s root cause processes.These Foxglove layouts make it easy to build a collection of ready-to-go dashboards for the whole company to use. Sharing a snapshot of a workspace for collaborating with another team member has now become as simple as copying and pasting a link.
Foxglove has proven to be a critical part of Waabi’s development toolchain. It has allowed the Waabi team to stay focused on building the innovative AI-powered software for its virtual driver and simulation platform, while having a solid visualization and data indexing platform supporting it.
Foxglove has enabled Waabi to better organize its driving data, allowing the team to inspect and tag that data, visualize both on-the-road and simulated behaviors, and debug autonomous driving performance. These features have helped Waabi to better visualize how well their latest features work, and use those findings to prioritize their development roadmap as they rapidly make progress in bringing safe and scalable self-driving truck technology to the road.