Laprit Luhar

Embedded systems and firmware engineer with 5+ years of end-to-end product development across industrial IoT, medical devices,

and consumer electronics. Take products from schematic and hardware design through firmware (C/C++, FreeRTOS, bare-metal),

Embedded Linux bring-up, and edge AI integration to production and field deployment. hardware and firmware lead for an AIenabled industrial weight measurement product line.

Embedded Systems & Firmware Engineer (Team Lead) | Soaire (So-Air) Apr 2024 – July 2026

• Lead hardware and firmware development of the WMS product line (v3.0–v4.7) — an AI-enabled Waste Management System

(V3) that evolved into the industrial Weight Measurement System (V4) — deployed at 12+ live customer sites; manage and mentor a team of 3 engineers.

WMS V3 to V4 Platform — Embedded Linux Firmware (Raspberry Pi / Radxa)

• Developed WMS V4 platform firmware on Raspberry Pi Zero 2W/5 and Radxa Rock 5C: weight stabilization, HDMI display HMI,

Wi-Fi manager, self-update OTA, and ToF sensor integration in multithreaded Python.

• Ported the product platform from Raspberry Pi to Radxa Rock 5C: performed board bring-up, resolved Wi-Fi driver, I2C, UART,

RTC, and camera-stack issues, and authored a Picamera2-equivalent camera/video library.

• Built an 11-stage automated first-boot provisioning pipeline (partition expansion, SSH/Wi-Fi setup, deploy keys, OS encryption)

validated on RPi and Radxa — cut machine deployment from 8 hours to 30 minutes.

WMS V4.0.8 — Edge AI Inference Pipeline

• Designed a multithreaded camera inference pipeline for edge AI with a custom thread scheduler; profiled RAM and per-thread performance on resource-constrained boards.

WMS807 Custom PCB & Power Hardware

• Root-caused weight instability to a power-rail brownout (shared supply sagging under RPi load, disrupting ESP32 + HX711) and redesigned the power distribution to eliminate it.
• Designed the custom WMS807 PCB with multi-rail power architecture, including component selection and KiCad library development; performed on-site installation, calibration, and customer training.

Low-Power Wireless Sensor Node

• Diagnosed and fixed a 2x sleep-current regression (15 µA to 7 µA) in coin-cell sensor firmware through duty-cycle and lowpower optimization.

IMS — Inventory Management System

Embedded Firmware Consultant — BottomLess IoT Pour-Measurement System | Independent / Client Project Jun

2023 – Mar 2024

Device Firmware & Power (nRF52 / Zephyr)

• Owned nRF52 (Zephyr/nRF Connect SDK) firmware for a coin-cell BLE liquid-measurement device: duty-cycle engineering,

flash/RTC event storage, BLE beacon link to ESP32 hub, and DFU/OTA workflows.

• Reduced sleep current from 240 µA to 15 µA and cut duty-cycle time from 9.5 s to 5.8 s (PPK-verified), extending modeled

CR2450 battery life to multi-year operation.

Sensing, Calibration Algorithms & TinyML

• Improved VL53L4CD ToF accuracy from ~10 mm to 1–2 mm error via structured characterization (range timing, sigma threshold,

averaging) and calibration algorithms (cubic interpolation, Kalman filter, lookup-table correction).

• Integrated TinyML models (Edge Impulse / TensorFlow, random forest) on nRF52833 for cap-fit and pour-event classification;

benchmarked ML vs rule-based detection accuracy.

• Contributed measurement-algorithm flowcharts to the product

Imaging Pipeline & Calibration Machines

• Built an ESP32-CAM imaging pipeline (capture, SD buffering, timestamped cloud upload) and designed & built two generations of automated calibration machines (mechanical + firmware + weight-scale integration).

Firmware Engineer | ElectrifyServices LLP Aug 2022 – May 2023

• Delivered firmware development and hardware debugging across five client product lines (IoT, consumer, industrial).
• BottomLess (IoT Pour Measurement): wrote register-level CH101 ultrasonic and LIS2DW12 drivers on nRF52 from datasheets,

fusing ultrasonic + accelerometer data for tilt/pour detection at ~240 µA sleep; ported firmware nRF52832 to nRF52833.

• DECT Headset: implemented the EAP serial protocol from specification (framing, checksum, supervisor/numbered frames) on

MAX32660 over UART/I2C under FreeRTOS, integrating with the RTX1090 DECT module.

• Smart Lock: built ESP32 firmware with capacitive-touch calibration, guest-key management, auto-lock logic, AES-128 encryption,

and motor PWM control; reduced idle power to 38 µA.

• DWS / Packing Eye: authored VL53L1X (C++ to C port) and HX711 libraries, implemented multi-sensor fusion and calibration on a custom Yocto Raspberry Pi image.
• Jetson Nano Carrier Board: performed bring-up, rework, power-fault isolation; prepared client-facing scoping documents,

quotations, and block diagrams for new projects.

R&D Head — Medical Devices | Motima International Pvt Ltd Oct 2021 – Jul 2022

• Led end-to-end product R&D — enclosure design, circuit development, embedded firmware, and production release — for a portfolio of medical devices.
• Oxygen Analyzer & Nebulizer: designed and released both devices from concept to production.
• Portable ECG Machine: built the device end-to-end including GUI design for a 3.5-inch resistive touch display.
• Oxygen Concentrator: designed the device interface circuit.

Technical Team Leader | Robocon (National Robotics Competition) Apr 2018 – Jul 2019

• Ranked 1st in India in Stage 1 of Robocon 2019 leading mechanical, firmware, and hardware teams; built a semi-automatic quadruped robot and a semi-automatic holonomic robot implementing inverse kinematics and PID control on SAM3X8E

(Cortex-M3) and ATmega2560

B.E. Mechanical Engineering — Saffrony Institute of Technology (CGPA 7.26) 2019