AAI Innovations GmbH
Tool Vendor / Service Provider
AAI Innovations GmbH is a Berlin-based software company founded in 2017, focused on developing advanced, data-driven solutions for the creation, simulation, and validation of complex systems. Formerly operating as Automotive Artificial Intelligence (AAI) GmbH, the company reflects its strategic repositioning with a broader, cross-industry focus.
While our origins lie in the automotive sector, our capabilities and applications extend across industries where scalable, high-fidelity simulation and intelligent data processing are required.
Our flagship platforms, RepliMap and ReplicaR, enable customers to build and edit HD maps, enrich 3D environments, and simulate sensor data with both realism and scalability. From rapid, object-level testing to physics-based sensor simulation (camera, LiDAR, radar), we support the full development lifecycle, from early-stage validation to large-scale scenario execution.
Beyond individual tools, our strength lies in combining deep engineering expertise with open standards such as ASAM OpenDRIVE, enabling seamless integration into existing development ecosystems. We provide scalable solutions in sensor modeling, scenario generation, and high-quality data workflows, supporting both domain-specific applications and cross-industry use cases.
With a growing international footprint and customers across Europe, Asia, and North America, AAI Innovations GmbH supports global development programs with platforms designed for engineering precision, scalability, and compliance with industry standards.
ASAM related products
RepliMap
Type
HD Map Authoring & OpenDRIVE Road Network Editor
Functionalities
RepliMap is a dedicated engineering platform for creating, editing, validating and maintaining simulation-ready road networks based on ASAM OpenDRIVE.
Unlike conventional GIS or CAD tools, RepliMap works directly with the OpenDRIVE road network structure, preserving geometry, lane topology and infrastructure information throughout the engineering workflow.
Key capabilities include:
Native OpenDRIVE editing of road geometry, lane sections, junctions, connectivity, elevation profiles, superelevation, road markings, traffic signals and infrastructure objects.
Import and conversion of geographic and HD map data, including GeoJSON and HERE HD Maps, with LiDAR point clouds, camera imagery and Mapbox data available as references.
Integrated validation of road geometry, lane connectivity, topology consistency and OpenDRIVE structure.
RepliCat, an integrated AI assistant that helps engineers explore road network data, understand modelling workflows and prepare map modifications through natural-language instructions.
Combined road network and 3D scene development, including buildings, vegetation and roadside infrastructure.
Export of ASAM OpenDRIVE (.xodr), FBX 3D assets with CARLA-specific options, and GeoJSON.
RepliMap supports ADAS validation, automated driving development, simulation engineering, HD map production and digital twin creation.
Supported ASAM Standards
ASAM OpenDRIVE®
ReplicaR
Type
Scenario-Based Simulation & Virtual Validation Platform
Functionalities
ReplicaR is a cloud-native simulation and virtual validation platform for ADAS and automated driving systems. It enables engineering teams to generate realistic driving scenarios, execute repeatable virtual tests and evaluate vehicle functions across a wide range of traffic conditions and Operational Design Domains (ODDs).
Key capabilities include:
Automated scenario generation with AutoScenario, using naturalistic traffic simulation, and MyScenario, using knowledge-based Morphosynthesis.
ASAM OpenDRIVE road network libraries and custom map imports with integrated MapChecker validation.
Open-loop and closed-loop vehicle simulation, including configurable traffic participants, driving behaviours and environmental conditions.
Fast object-list simulation and physics-based sensor simulation for camera, LiDAR and radar.
Synthetic sensor data generation with SynthData and SenVironment, including ASAM OSI recordings and ROS 2 bags.
Automated test suites, scenario extraction, criticality analysis, regression testing and simulation analytics.
Integration with existing development and simulation toolchains through co-simulation interfaces and the ReplicaR Connect Python SDK.
ReplicaR supports scalable simulation campaigns in AAI-managed cloud environments or customer-hosted cloud infrastructure.
Supported ASAM Standards
ASAM OpenDRIVE®
,
ASAM OSI®
OpenDRIVE Map Validator
Type
Automated OpenDRIVE Validation Service (MCP)
Functionalities
The OpenDRIVE Map Validator enables engineers to check ASAM OpenDRIVE road networks for structural quality and simulation readiness directly through compatible AI assistants using the Model Context Protocol (MCP).
The service connects conversational AI workflows with deterministic engineering validation, combining automated quality checks and simulation-based verification without modifying the original road network.
Key capabilities include:
Automated ASAM OpenDRIVE quality checks for structural and logical consistency.
Additional checks for road continuity, topology and simulation-related issues.
Traffic simulation to assess whether the submitted road network can be driven in the validation environment.
Consolidated validation results, including a driveability verdict and detailed checker logs.
Read-only processing that preserves the original uploaded OpenDRIVE file.
Direct execution through compatible MCP clients, including Claude, using an AAI customer account.
The service helps developers identify map issues before using road networks in scenario generation, virtual testing or downstream simulation environments.
Supported ASAM Standards
ASAM OpenDRIVE®
AutoScenario Driving Scenarios
Type
Automated Driving Scenario Generation Service (MCP)
Functionalities
AutoScenario generates simulation-ready driving scenarios from ASAM OpenDRIVE road networks. Accessible through compatible AI assistants using the Model Context Protocol (MCP), the service makes automated traffic scenario generation available without requiring direct access to a full simulation platform.
Powered by ReplicaR's AutoScenario engine, it uses traffic simulation to identify driving situations and produce corresponding ASAM OpenSCENARIO XML files.
Key capabilities include:
Automated traffic simulation using customer-provided ASAM OpenDRIVE road networks.
Identification and extraction of driving events and traffic interactions.
Generation of individual ASAM OpenSCENARIO XML (.xosc) files together with the corresponding OpenDRIVE (.xodr) map.
Configuration of scenario duration, traffic density, vehicle placement, traffic participant types and time of day.
Vehicle-under-test configuration using predefined vehicle profiles, with adjustable sensor range.
Direct execution and retrieval of generated scenario files through compatible MCP clients, including Claude.
AutoScenario supports scenario-based testing, simulation preparation and automated driving development by simplifying the creation of reproducible traffic scenarios.
Supported ASAM Standards
ASAM OpenDRIVE®
,
ASAM OpenSCENARIO® XML
HD Mapping & Digital Twin Engineering
Type
HD Map Production & Digital Twin Engineering Services
Functionalities
AAI provides engineering services for creating high-definition road networks and digital representations of real-world environments for automotive simulation, ADAS validation, automated driving development and infrastructure applications.
Combining automated data processing with specialised engineering tools and quality assurance, AAI transforms geographic and sensor data into structured road models and reusable simulation environments.
Key capabilities include:
Road and infrastructure data acquisition using mobile mapping, camera imagery, LiDAR and existing geographic datasets.
HD road network modelling covering road geometry, lane topology, junctions, elevation, traffic infrastructure and semantic attributes.
Conversion of geographic and mapping data into ASAM OpenDRIVE, including engineering review and quality assurance.
Reconstruction of existing roads, urban environments, test tracks, proving grounds and planned infrastructure.
Three-dimensional environment creation, including buildings, vegetation and roadside objects.
Creation of reusable digital road models and simulation assets using RepliMap, with optional integration into ReplicaR and customer simulation toolchains.
Project-specific delivery of OpenDRIVE road networks and corresponding 3D assets.
Services can cover individual engineering tasks or the complete workflow, including data collection, network modelling, validation and integration into the customer's simulation environment.
Supported ASAM Standards
ASAM OpenDRIVE®
Data Annotation & Labelling Services
Type
Automotive Data Annotation & Labelling Services
Functionalities
AAI provides scalable data annotation and labelling services for automotive AI development, perception systems, driver assistance functions and automated driving applications.
Combining automated annotation technologies with human quality assurance, AAI supports the creation of structured datasets for AI training, perception validation and benchmarking.
Key capabilities include:
Manual, semi-automated and automated annotation of images, video sequences and sensor data.
Object detection and classification, semantic segmentation, road and lane annotation, and automotive scene interpretation.
Human-in-the-loop workflows combining automated pre-labelling with manual review and correction.
Integration of annotation workflows into customer-provided labelling platforms and data environments.
Development of specialised annotation tools and workflows for project-specific requirements.
Quality assurance processes with defined acceptance criteria, consistency checks and performance monitoring.
Scalable data processing for large automotive datasets and recurring production programmes.
Services are tailored to customer-specific annotation requirements, data formats and quality targets, supporting both development datasets and validation-oriented ground truth production.
Keywords
ASAM OpenDRIVE, ASAM OpenSCENARIO, ASAM OSI, HD Mapping, HD Map Authoring, Digital Twins, Road Network Modelling, Scenario Generation, Traffic Simulation, ADAS Validation, Automated Driving, Virtual Validation, Sensor Simulation, Synthetic Data Generation, OpenDRIVE Validation, Data Annotation, Automated Labelling, Ground Truth, AI-Assisted Engineering, Model Context Protocol (MCP), Simulation Automation, RepliMap, ReplicaR, RepliCat, AutoScenario
