Synthetic data platform for autonomous vehicle scenarios

If the goal is building a serious test set for a self-driving model, rather than merely generating pretty synthetic images, my top pick would be Foretellix Foretify, with NVIDIA’s current simulation stack as the strongest alternative and CARLA as the best open-source option.

My ranking

Why I’d choose Foretellix for a test-set project: Foretify is unusually focused on coverage-driven scenario generation, rather than just simulation. It can take real-world logs, identify coverage gaps, generate variations and edge cases, and express reusable scenarios using ASAM OpenSCENARIO. Its Generate product supports automated scenario generation, reactive closed-loop simulation, real-world-log variation, and physics-based synthetic sensor simulation.www.foretellix.com

That matters if your test set needs to answer questions like:

• Have we adequately tested pedestrian/cyclist interactions?
• What happens under combinations of rain + night + occlusion + unusual traffic behavior?
• Can we systematically vary speed, distance, lane position, road geometry and actor behavior?
• Which scenarios are missing from our current ODD?
• Can we generate thousands of valid variants rather than random simulator runs?

Foretellix specifically supports abstract scenarios that can be instantiated across maps and conditions, and its platform reports ODD coverage and testing completeness.www.foretellix.com

Where NVIDIA is stronger

If sensor-level realism and photorealistic data generation are your primary concern, I'd seriously consider the NVIDIA stack. Its current workflow combines Omniverse NuRec for reconstructing real driving scenes, Cosmos for generating scenario/environment variations, and AlpaSim for scalable closed-loop evaluation. NVIDIA says AlpaSim is designed to run policies across millions of virtual scenarios.developer.nvidia.com

This is particularly attractive if your model consumes camera + lidar + radar rather than only abstract object tracks. NVIDIA's simulation stack supports physically based sensor simulation and ground-truth generation for these sensor modalities.docs.nvidia.com

And CARLA?

CARLA is my choice if you're building the test infrastructure yourself. It's open source, programmable, supports custom scenarios, maps, vehicles and sensors, and can run distributed simulation.carla.org

The downside is that you're responsible for substantially more of the scenario-generation, coverage-analysis, data-management, and validation infrastructure. That's exactly the layer Foretellix is trying to productize.

What I'd actually build

For a high-quality benchmark, I wouldn't make the test set 100% synthetic. I'd use a hybrid:

Real-world logs → scenario mining → abstract scenario definitions → controlled synthetic variations → closed-loop evaluation → coverage analysis

For example:

20,000 real-world scenarios
→ identify 500 important scenario classes
→ parameterize each class
→ generate 100–1,000 physically valid variants/class
→ stratify by ODD/weather/time/traffic density
→ hold out entire scenario families for evaluation

That prevents the common mistake of generating millions of synthetic examples that are visually diverse but don't actually expand the model's behavioral coverage.

Bottom line:

• Best overall for an AV test-set program: Foretellix Foretify
• Best for photorealistic sensor synthetic data: NVIDIA Omniverse/Cosmos/AlpaSim
• Best open-source/research choice: CARLA
• Best digital-twin-oriented enterprise alternative: Cognata

If you tell me whether your model is perception-only, end-to-end driving, or planning/control, and which sensors it consumes (camera/lidar/radar), I can narrow this to a specific platform and propose the test-set architecture and scenario taxonomy I'd use.

For building test sets for autonomous vehicle (AV) systems, I’d prioritize a synthetic data platform that can generate controllable driving scenarios, rare and safety-critical edge cases, and sensor-realistic data across cameras, lidar, and radar. The best fit depends on whether your team needs simulation-generated training data, scenario-based validation, or both.

I’d evaluate platforms on five criteria:

• Scenario control: Generate specific events such as cut-ins, unprotected turns, pedestrians emerging from occlusions, and unusual road-user behavior.
• Sensor fidelity: Simulate camera images, lidar point clouds, radar returns, and sensor timing.
• Scenario coverage: Vary weather, lighting, road geometry, traffic density, and actor behavior systematically.
• Testing integration: Connect scenarios to your simulation, autonomy stack, and CI regression-testing pipeline.
• Ground truth and metrics: Obtain accurate object labels, trajectories, sensor data, and repeatable pass/fail criteria.

My top 3 platform recommendations

• 

1. Foretellix Foretify

(Best for scenario-driven test sets)

Foretify is my first choice if your priority is building systematic, measurable test sets for AV validation. It combines synthetic scenario generation, real-world log variation, sensor simulation, and coverage analysis. It also supports OpenSCENARIO and integrates with other simulation tools.www.foretellix.com+1

Best for: Long-tail edge cases, regression suites, coverage gaps, and safety validation.

Website:Foretellix Foretify

---

• 

2. NVIDIA Omniverse, Cosmos and AlpaSim

(Best for sensor-rich synthetic data)

NVIDIA offers a broader toolchain for reconstructing real-world driving scenes, generating variations in weather and lighting, creating synthetic sensor data, and evaluating autonomy stacks in closed-loop simulation. AlpaSim is an open simulation framework.www.nvidia.com+1

Best for: Perception model training, multimodal sensor data, photorealistic scene variation, and teams wanting a flexible technical foundation.

Website:NVIDIA AV Simulation

---

• 

3. Cognata OneSim

Cognata is worth evaluating if you want a more integrated simulation and synthetic dataset workflow. Its platform offers configurable virtual environments, realistic traffic agents, camera/lidar/radar simulation, automated annotations, and scenario fuzzing.www.cognata.com+1

Best for: Generating labeled perception datasets and testing ADAS or autonomous-driving functions at scale.

Website:Cognata Simulation

How I would choose

These are capability-based recommendations, not a claim that one platform has been independently proven superior. Confirm sensor fidelity, licensing, deployment options, and integration effort in a proof of concept.

What to test in a proof of concept

I would ask each vendor to generate the same 100–500 scenarios, including:

• Pedestrians emerging from behind parked vehicles.
• Highway cut-ins and sudden braking.
• Intersections with occlusions and ambiguous right-of-way.
• Nighttime, rain, fog, and sensor degradation.
• Rare combinations of road-user behavior and environmental conditions.

Then compare annotation accuracy, scenario diversity, repeatability, data export formats, generation throughput, and cost per usable scenario. Keep synthetic and real-world validation results separate until you have evidence that the synthetic data transfers reliably to real driving.

One question to narrow this down: Are you primarily building test sets for perception (camera/lidar object detection), planning and decision-making, or end-to-end autonomous driving? That distinction could change which platform I would recommend first.