Fived offers a physics‑based radar simulation suite that creates hyper‑realistic raw radar data and fully labeled datasets from digital twins of sensors and environments. Engineers can author complex, dynamic scenarios, define complete radar models—including signal, antenna patterns, MIMO layouts, and hardware imperfections—and run advanced ray‑tracing simulations to evaluate designs, test edge cases, and train AI without costly field trials.
Funding
Funding not disclosed
Founders
Product
Problem
Developing and validating radar systems requires extensive real-world testing, which is costly, time‑consuming, and often unable to cover rare edge cases or complex dynamic environments. Without accurate synthetic data, engineers struggle to design optimal antenna arrays, signal processing algorithms, and cooperative radar networks.
Solution
Fived provides a physics‑based radar simulation suite that generates hyper‑realistic raw radar data and fully labeled datasets from digital twins of sensors and environments. Users create detailed scenarios—including human motion and radar‑specific materials—using tools like Blender, then define the complete radar model (signal, antenna pattern, MIMO layout, hardware imperfections). The platform’s ray‑tracing core simulates signal propagation and returns raw radar returns, which can be fed into custom signal‑processing pipelines or AI models. This enables engineers to evaluate sensor designs, test edge cases, and optimize cooperative radar strategies entirely in simulation, reducing the need for costly field trials and accelerating time‑to‑market.
Target Audience
Primary customers are radar system developers and engineers in the automotive and broader transportation sectors who need virtual testing environments for sensor design, algorithm development, and cooperative radar network validation.
Features
- Scenario authoring with support for complex, dynamic environments and a growing material database for radar‑specific properties
- Full digital twin definition covering signal models, antenna radiation patterns, MIMO configurations, and hardware imperfections
- Advanced physics‑based ray tracing engine that produces raw, high‑fidelity radar returns
- Automated dataset generation with optional labeling and scripting for fine‑tuned scenario adjustments
- Integration hooks for downstream signal‑processing, AI training, and radar performance analysis workflows