Subhraneil Das

Designed and implemented a PyTorch-based system for real-time indoor trajectory estimation through sensor fusion of raw Inertial Measurement Unit data, without reliance on GPS. Reduced trajectory prediction error for ambiguous human walking cases by ~4%. Achieved prediction accuracy within the benchmark standard of ±6m, addressing common sources of measurement drift and bias. Verified model reliability through testing on varied human motion profiles to cover corner cases.

Designed and implemented a flow cytometry-based sensor data acquisition system for photovoltaic hardware, covering signal capture, conditioning, and logging pipelines to detect pathogens in fluorophore-stained samples. Employed photonics and design space exploration against 4 fluorophore samples varying between pure and mixed concentrations at respective excitation wavelengths for system feasibility. Successfully replaced an XIMEA CMOS image sensor with a PIN photodetector for fluorophore detection.

Completed multiple project use-cases focused on embedded firmware and sensor systems with strict adherence to functional safety and quality standards. Designed and executed C-based verification tests for the AUTOSAR Watchdog Manager module in compliance with ISO 26262 and MISRA C, attaining 74% structural code coverage.

Developed and validated a Velocity Matching (VM) Transfer Alignment (TA) algorithm for an Inertial Navigation System using model-based design in Simulink. Estimated nine dynamic states and characterised measurement and alignment errors for system calibration. Verified algorithm stability through 15 successful simulation runs and converted the validated model into deployable C firmware for an indigenous INS platform.

Specialised coursework: Embedded Computer Architectures, Real Time Software Development, Real Time Systems, Advanced Computer Vision & Pattern Recognition, Pervasive Computing, Design of Digital Systems. Research Focus: Application of deep learning in inertial odometry, data acquisition from GNSS and inertial sensors.

Coursework: Analog Electronics, Digital Logic Design, System Design on VERILOG, Analog & Digital Communication. Minor Specialisation: Embedded Networking, Real Time Systems, System Design on VHDL, Embedded Firmware Development.