The startup provides a unified workflow for searching and analyzing multimodal datasets, including camera, radar, and lidar data, to identify trends and data gaps. Their tools enable teams to visualize scenarios and automate safety validation, enhancing the development of spatial intelligence in various applications.
Funding
$9.3M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
Developing spatial intelligence applications, such as autonomous vehicles and robotics, requires analyzing vast multimodal datasets (camera, radar, lidar, etc.) to identify trends and potential safety issues. This process is often hampered by the lack of unified tools for searching, visualizing, and understanding these complex datasets, leading to inefficiencies and potential oversights.
Solution
The company provides a platform for unifying multimodal datasets, enabling teams to efficiently search, visualize, and analyze data from various sensors. Their tools allow users to uncover trends, identify data gaps, and automate safety validation through scenario-based testing. By transforming episode logs into vectorized scenes, the platform facilitates the computation of competency and safety metrics at scale, accelerating the development and deployment of spatial intelligence applications. The platform includes open-source tools for unifying multimodal log data and cutting-edge embodied AI research.
Target Audience
The primary target audience includes embodied AI teams in automotive, agriculture, industrial automation, and robotics, as well as researchers working on spatial intelligence, cross-modal learning, and multimodal datasets.
Features
- Unified workflow for ingesting and managing multimodal datasets (camera, radar, lidar, audio, controls) in MCAP format
- Advanced search capabilities to identify specific scenarios based on dataset, lane count, lighting conditions, object interactions, and other parameters
- Dataset metrics and trend analysis to uncover data gaps and curate balanced training sets
- Scenario-based testing to estimate key safety metrics and accelerate deployment
- Open-source spatial intelligence models and source code for customization and research
- Tools for morphing real episode logs into vectorized scenes for automated task detection
- Open Scenario Map: a catalog of annotated scenarios sourced at scale, designed to be unbiased and scalable