Advanced Driver Assistance System

HL Klemove provides the complete Sense-Think-Act solutions for the
Advanced Driver Assistance (ADAS) and Automated Driving (AD)
systems in today’s automobiles.

AI Software

Our experienced team of engineers has exposure on end-to-end implementations, from sensors to the vehicle CAN interface, down to the safety-critical ECUs which control vehicle motion and navigation. At our R&D Center in India, our in-house team focuses on four key areas of Artificial Intelligence (AI) to maintain this momentum.


Data

The effectiveness of any AI system is directly dependent on the quality of the data used in its construction. At our company, we offer the expertise for creation of high-quality and high-value datasets to drive today’s AI algorithms.


Our expertise:
  • creation and curation
  • Data augmentation using synthetic information
  • Data annotation
  • Automation tool development
  • Analytics tool Development


Algorithm

Artificial Intelligence (AI) has disrupted the automotive industry. Neural networks have emerged as the key differentiating feature of modern ADAS and AD systems. As an early adopter of the technology, HL Klemove has been at the forefront of AI research and development. Our group designs and develops Deep Learning (DL) neural networks for high-level safety-critical applications such as 2D and 3D Object Detection, Semantic Segmentation, and Road Feature Detection. These optimized DL solutions are capable of accepting inputs from multi-modal sensors such as camera, radar, LiDAR, and GPS, and are compatible with industry-standard hardware platforms such as Nvidia, Qualcomm, and Texas Instruments, among others.


We specialize in solutions for:
  • Design and development of DL neural networks
  • Neural Network porting and integration for edge devices
  • Neural Network migration and optimization
  • Development of multi-modal Sensor Fusion neural networks
  • Development of platform-agnostic neural networks


Platform

For more than a decade, HL Klemove has led the automotive industry into research and development of the most advanced solutions which rely on conventional Computer Vision (CV) and Machine Learning (ML) techniques. Apart from the development of heuristic algorithms for use-cases such as scene perception and object detection, we have demonstrated competency in sensor calibration, sensor information processing, integration of vehicle dynamics, output signaling, and integration of CV/ML and DL network solutions.


Our expertise:
  • Computer Vision(CV) and Machine Learning(ML) algorithm development
  • Design and development of ADAS/AD applications for specific use-cases
  • Development of scene refinement solutions like object tracking
  • Integration and application of vehicle dynamics
  • Sensor Fusion for sensors such as camera, radar, ultrasonic, LiDAR, GPS
  • Development of V&V tools


Embedded

ADAS and AD solutions are safety-critical and demand real-time performance. Construction of any such real-world solution also requires experience in the construction and implementation of edge systems and devices capable of addressing the functional and performance requirements in the automotive environment. Our team has an established practice of edge implementations of real-time ADAS/AD solutions on platforms provided by leading semiconductor manufacturers such as Nvidia, Qualcomm, and Texas Instruments. We have significant exposure to ADAS/AD software development for evaluation as well as production-ready hardware, with multiple sensors, and signal integration for real-world vehicles.


Our expertise:
  • Design, development and testing of ADAS/AD application software(ASW)
  • Migration, porting and integration of conventional CV/ML and DL algorithms
  • Optimization of algorithms and applications for CPU, GPU, Deep Learning, or other dedicated hardware accelerators
  • Multi-core automotive ASW for cutting-edge SoCs
  • Integration of sensors for automotive edge platforms