Blog post
The Way of Data: How Sensor Fusion and Data Compression Empower Autonomous Driving
ADAS and Autonomous Driving Development
As an AUTOSAR Associate Partner, Intellias helps Tier 1 automotive companies and OEMs design AUTOSAR-based architectures for ADAS/AD (Autonomous Driving) solutions, achieve industry compliance, and align with functional safety standards to meet their goals for developing autonomous driving software at scale.
Our ADAS development services empower clients with modern AI/ML technologies, cloud platforms, highly secure AUTOSAR solutions, and real-time operating systems (RTOS). We cover full-cycle ADAS/AD solutions development, from sensor programming and embedded OS integration to real-time data processing, algorithm optimization, and development of ADAS applications that automate critical driving functions.
Trusted by leading companies worldwide
ADAS and Autonomous Driving development competencies
AI and machine learning
We have experience with market-leading AI/ML frameworks, including TensorFlow, Keras, Qualcomm SNPE, Caffee, and cuDNN. We prepare extensive datasets with data labeling, validation, and augmentation for AI networks supervised learning using linear regression, decision trees, case-based reasoning, and Kernel estimators, as well as reinforcement learning applying Q-learning algorithms, Boltzmann machines, Deep Q-Networks, and inverse learning models.
AUTOSAR for ADAS/AD
Intellias is an AUTOSAR Associate Partner, specializing in AUTOSAR Classic and Adaptive Platforms integration and migration to RTOS and AUTOSAR-based solutions. We develop complex and secure architectures based on AUTOSAR standards, integrating third-party components and configuring and validating various AUTOSAR components, including Application Software and Bootloader.
ADAS solution design
Intellias designs ADAS/AD architectures from scratch based on a client’s needs, analyzing the feasibility of initial ideas to determine how they match modern technological capabilities and providing an insight-based optimization strategy, followed by PoC design and implementation.
ADAS solution development
Our experts know how to build custom ADAS/AD solutions aligned with functional safety standards. We implement AI/ML technology specifically for clients’ embedded platforms, improving data processing and verification capabilities with extended data sets required for automated driving features.
ADAS solution optimization
We provide operational support and infrastructure setup, identify infrastructure performance KPIs, configure the environment and infrastructure for ADAS/AD solutions. Our experts optimize data storage access, data caching algorithms, and data preloading, as well as execute test suites to ensure data processing quality meets required KPIs and gather critical metrics on processed data.
ADAS solution integration
We assist in integrating ADAS map solutions into navigation content products, including Minimal Risk Maneuver, Electronic Horizon, Advanced Driver Assistance Systems, and Improved Fuel Consumption. By developing and integrating sensor drivers, we fill ADAS data processing pipelines with safety-critical data.
Automotive OS capabilities
Our automotive OS integration capabilities extend from low-level programming, sensor firmware and OS drivers’ development, profiling, and testing to middleware integration of AI/ML frameworks into a board support package (BSP). We cover development of apps with AD features, integration with Linux and RTOS, acceleration of hardware and integration with SDKs such as proprietary ADAS, CV, Kanzi, GL Studio, Qt, and Android Automotive.
Validation and verification
We have a deep understanding of full-cycle development, integration, and validation of AUTOSAR-based ECU components and ADAS/AD safety functions to ensure industry compliance and alignment with requirements. Our engineering teams perform functional, integration, and qualification testing, validating and verifying ISO/SAE 21434 criteria and simulating ADAS/AD algorithms using modern techniques and proprietary frameworks.
Software and hardware testing
We develop and perform unit, function, integration, and qualification testing using manual/automated tests and applying Model in Loop (MIL) and Hardware in Loop (HIL) testing models. We also perform ADAS/AD penetration testing to ensure the development of highly reliable autonomous driving solutions that can meet security requirements and industry standards and handle real-life road circumstances.
Compliance and certification
We support our clients with validating ADAS features using the NDS Certification Bench and implement alignment strategies to comply with industry standards for safety and security, including AUTOSAR, A-Spice, and FuSa/ISO 26262 implementation requirements. In engineering ADAS/AD solutions, we follow AUTOSAR and ADASIS requirements for seamless integration with custom software.
Partnerships and certifications
Navigating change. Our insights
How can we help you?
Get in touch with us. We'd love to hear from you.
