Tools for vehicle intelligence

Develop and validate next-generation AD and ADAS, and general vehicle software. Move from design to start of production (SOP) up to 4x faster and release with confidence using state-of-the-art AI.

Contact us

“Applied Intuition works hand in hand with our engineering teams to integrate their tools with our quite complex codebase.”

Nikos Michalakis

VP of Software Platform, Toyota

“We chose to work with Applied Intuition because of its state-of-the-art simulation technology. We are comprehensively validating our system for various complex situations that could arise in the real world.”

Takao Asami

Senior Vice President, Nissan

“The Applied Intuition team has demonstrated their expertise and has equipped us with tools to accelerate the safe development of commercial trucks in a financially viable way.”

Michael Fleming

Founder and Board Member, Torc Robotics

“Applied Intuition’s solution allows us to highly automate our scenario-based, data-driven engineering workflows and adapt jointly developed applications as white-box solutions for the overall management of high-performance, safety-critical AD systems.”

Gero Kempf

Executive Vice President for ADAS/AD, Audi

Challenge

ADAS and AD increasingly leverage end-to-end (E2E) architectures that require new approaches to data, machine learning (ML), and simulation. In-vehicle experiences spanning infotainment, mobile applications, and body controls are also becoming more software-defined. Automakers must develop next-generation features while accelerating development and complying with regulations such as UN R157 and UN 171.

Simulation environments are manually created and low fidelity

Significant investments are required to test all vehicle software and meet all regulations

Insufficient data and ML to efficiently scale programs

Extended development timelines lead to delays and potential impact on safety

Applied Intuition’s approach

Applied Intuition’s toolchain provides large-scale data and ML infrastructure, neural simulation, AI agents, and full vehicle testing. Engineering teams can develop, validate, and release next-generation E2E AD and ADAS up to 4x faster, with confidence.

AI-powered simulators use technologies such as neural rendering, diffusion, and world models to achieve realism and scale

AI agents bring power-user capabilities to every user, performing complex tasks up to 10x faster

Data and ML infrastructure helps engineers curate over 100 PB of data and 1 million diverse scenarios

Full-vehicle software testing, type approval, and in-service monitoring solutions help comply with regulations such as UN R157 and UN 171

The vehicle intelligence development cycle

Data Engine

Collect and ingest 100s of PBs of drive data. Leverage large-scale data infrastructure to automatically filter, tag, and label logs. Additionally, augment with synthetic scenarios using fuzzing, sensor diffusion, and world model samples. Curate a realistic, diverse library of 1M+ scenarios.

Development

Train 1- and 2-stage E2E models on scale- and cost-optimized training and inference platforms. For 1-stage systems, train end-to-end. For 2-stage systems, train perception and behavior models. In all cases, leverage open- and closed-loop simulators and evaluation metrics. Use the large-scale scenario library for all development.

Validation

Obtain type approval and release with confidence. Validate ADAS/AD using different simulation environments, including closed-loop sensor-level Neural Sim. Link all results to operational design domains (ODDs) and requirements for full system traceability. Automatically generate release and safety reports for type approval. Use the large-scale scenario library for all validation.

Explore the products

Simulation products

Neural Sim

AI-powered simulator that turns drive logs into interactive worlds for closed-loop ADAS testing and training

Dynamic scene reconstruction from drive logs

Camera, lidar, and radar simulation from novel views

Scenario generation diversifies and expands coverage to capture rare and edge-case scenarios

Faster than real-time rendering for high-throughput closed-loop training and testing

Learn more

Object Sim

Simulator for planning, prediction, and controls development

ISO 26262-certified simulator with a rich library of behaviors for vehicles, pedestrians, and other agents for realistic object simulation

Flexible, composable library of criteria to determine if a scenario passes or fails

Intuitive and easy-to-use workflows for creating, editing, and managing scenarios at scale

Learn more

Sensor Sim

Physically accurate simulator to develop, test, and validate sensor suites and perception systems

Library of validated sensor models to represent specific sensor hardware

Procedurally generated 3D worlds customized to an ODD

Programmatically generated ground truth labels ready for ML systems, such as bounding boxes, depth, semantic segmentation, and optical flow

Learn more

Log Sim

Log-based testing tool to identify, reproduce, debug, and resolve issues from real-world testing

Deterministic log data re-simulation to reproduce real-world behavior

Scenario extraction to generate complex synthetic scenarios from vehicle data

Performant web-based playback enabling accurate debugging and root-cause analysis

Learn more

Cloud Engine

Execution engine that runs object, sensor, and log simulations at scale in the cloud

Scalable, reliable, and cost-efficient infrastructure for large-scale parallel simulation and integration into continuous integration (CI) pipelines

Cloud-based workflows for simulation at scale, giving the ability to manage scenarios and test libraries, view results, and assess performance collaboratively across an organization

Interoperable and configurable infrastructure to run on AWS, Azure, Google Cloud, and on-premise

Learn more

VehicleSim

Comprehensive suite of software tools that enables realistic and dynamic simulation of vehicle behavior

Physically accurate vehicle models designed for use by novices and experts alike

High-performance computing (HPC) for increased coverage, locally and in the cloud

Real-time models ideal for software-in-the-loop (SIL), driver-in-the-loop (DIL), and hardware-in-the-loop (HIL) development

Learn more

HIL Sim

Hardware-in-the-loop simulator to identify safety-critical failures early and reduce development time

Object-level HIL testing for planning and controls development

Sensor-level HIL testing to validate a system from sensor input to actuation

Vendor-agnostic approach, supporting all mainstream hardware vendors

Learn more

Data products

Data Explorer

Visualization and query tool for collected log data

Cloud-based visualizer to search across logs, root-cause events, and export data

Rule-based ingestion and intelligent models to enrich data with relevant metadata, especially for unlabeled logs

Scalable querying engine to identify important logs

Learn more

Synthetic Datasets

Synthetic dataset generation tool optimized to train machine learning (ML) models

High-level dataset definition language and visual editor to easily define needed data

Dataset management tooling to view statistics, filter, and export data

Generated datasets are proven to improve model performance in published case studies

Learn more

Validation Toolset

Verification and validation tool to accelerate production deployments safely

Integrate test management artifacts and define an ODD to analyze system performance, measure safety, and improve test coverage

Support for abstract and logical scenario languages such as OpenSCENARIO V2.0 to easily generate scenario variations

Intelligently sample scenario variations to target failure regions and reduce computational execution cost

Learn more

Test Suites

Pre-defined scenario suites to comply with regulations and build a safety case

High-quality base scenarios that fit unique requirements and assess performance

Simulate regulatory testing protocols such as Euro NCAP, NHTSA, and more

Complex actor behaviors, routes, trigger conditions, and observers to run simulation tests out-of-the-box

Learn more

Map Toolset

Tool to create maps for use in simulation or on vehicles and generate test cases at scale

Web-based interface to create maps, edit maps, and validate maps

Parameter sweeps over different map attributes to generate scenarios at scale around points of interest

Generate realistic 3D worlds with minimal effort for a target ODD

Learn more

Applied Intuition Copilot

A generative AI-powered, prompt-based interface to interact with Applied Intuition’s ADAS and AD development platform

Generate simulation scenarios from simple text prompts

Create and edit maps using natural language prompts

Use natural language to generate SQL prompts

Tag and annotate drive log data automatically for search

Rapidly onboard and learn workflows in Applied Intuition’s ADAS and AD development platform

Learn more

Benefits

10x faster developer workflows

Automate manual tasks with AI agents. Reconstruct 3D environments with neural simulation pipelines. Complete tasks in minutes and hours, not days and weeks.

4x faster time-to-market

Utilize one unified platform for development, validation, and deployment. Reduce program timelines from an average of four years to 12 months with AI-powered tools.

1M+ curated scenarios

Leverage large-scale data infrastructure to automatically filter, tag, and label logs. Augment with synthetic scenarios using fuzzing, sensor diffusion, and world model samples. Output a diverse, realistic scenario library.

0 recalls

Leverage full vehicle software integration testing for 100% coverage of all virtualized software. Data and ML infrastructure enable collecting, root-causing, and fixing all issues from test fleets.

Frequently asked questions

Get started with the tools for vehicle intelligence

Accelerate the development and validation of next-generation ADAS, AD, and vehicle software.

Contact us