Sanglap Sarkar

A. Radar Tracking:

• Development of radar target tracking algorithms (detection post-processing, shape estimation, probabilistic data association, segmentation, k-nearest neighbor algorithm, interacting multiple model, extended Kalman filter, Kalman-based heading estimation, mirror/ghost detection, etc.).
• Responsible for the technical management of all radar tracking algorithm-related topics (base requirements and customer issues, both) of generations 5 and 6 radars.

B. Radar Perception:

• Radar perception algorithm development and technical management for (a) road border estimation using clothoid-based models, (b) trailer merge assist functionality, and (c) object height estimator modules.

C. Multi-Sensor Fusion:

• Multi-sensor fusion architecture and algorithms with long- and short-range radars, cameras, LiDAR, etc.
• Short-range, corner-mounted radar fusion/stitching for blind zone monitoring, improved object tracking, and enhanced classification [IP filling in prog].
• Master-slave-based 4-corner mounted short-range radar (SRR) and 2 forward-looking long-range (ARS) radar fusion (object kinematics and attribute fusion) in ADCU.

D. Radar-based ML:

• Rain Observer (rain, grass road, dry road, etc.) With radar detection properties (e.g., RCS, SNR) using Convolutional Neural Networks (CNN), multi-layer perceptron, random forest classifier, XGBoost classifier, and logistic regression-based supervised learning models [Internship Guidance] [PoC for next-generation radar].
• Supervised ML-based radar detection classification as ghost and real using logistic regression.
• Supervised ML-based target type classification (e.g., car, bike, pedestrian, truck, etc.). Using logistic regression.
• Unsupervised ML-based target heading estimation using Principal Component Analysis, Hough transformation, and IP filling in progress.
• Agentic AI-based radar business requirement processing using the CrewAI framework.

E. Radar Architecture:

• Architecture and interface finalization for multiple components (raw detection, radar tracking, perception, radar-based fusion, etc.). For Generation 6 radar.
• Project Management Topics:
• Product Owner role for the 'Radar Detection Processing and Tracking' team for the Daimler and Toyota projects.
• People Management Topics:
• Leading a team of 15.
• Performance appreciation and improvement discussions with team members.
• Addressing the needs of team members during on-site visits (China, Germany).

Products worked on:

• Long Range Radar : ARS620, ARS540, ARS510, ARS441, ARS512
• Short Range Radar : SRR630, SRR620, SRR510, SRR310, SRR62W
• Camera: MFC4xx

ADAS / ARAS Functions supported:

• Long Range Radar : ACC, EBA, FCW, US/China/Euro-NCAP
• Short Range Radar : BSD, LCA, OSE, RCTM, ToS, SoT
• Short Range Radar on Two-wheeler: ACC, BSM - (Advanced Rider Assistance System- ARAS)
• Sensor Fusion Project: Blind Spot Information System (BSIS)
• Tractor safety applications

Customers:

• ADAS (4-wheeler): Toyota, Subaru, General Motor, Ford, SGM China, Nissan, Volkswagen, Mercedes, BMW, Gelly, Daimler, Caterpillar
• ARAS (2-wheeler bike): Kawasaki, UltraViolet, BMW

Roles and Positions:

• Lead Architect and L2 expert. Nominated as one among seven L2 experts in Continental (globally) in Radar team of 1000+ on "Multi-Sensor Fusion Architecture". [Jul 2022 - Present]
• Lead Architect, L1 expert and Product Owner [Jan 2022 - Jul 2022]
• Senior Technical Architect, L1 expert, Scrum Master [Apr 2021 - Jan 2022]
• Senior Technology Specialist and L1 expert Selected as one among ten L1 experts in Continental (globally) in Radar team of 1000+ on "Radar Target Tracking". [Jul 2019 - Apr 2021]
• Technology Specialist [Jul 2018 - Jul 2019]
• Senior Executive Engineer [Jan 2017 - Jul 2018]
• Executive Engineer [Mar 2016 - Jan2017]

• Role: Algorithm Development and R&D
• Area: Machine Learning, Natural Language Processing (NLP), Agent Based Modelling (ABM), Deep Learning
• Contribution: Two international publications on Deep Learning and NLP [CIMSim 2015, Malaysia]

• “iSAP - Integrated Server Application Protocol” team using JAVA.
• Software Development at “Business Intelligence” team using C++.