Platform: dsPIC33 series microcontroller
Role: Firmware Developer
Status: 🟢 Deployed & under observation

Redesigned and redeveloped firmware for a multi-table solar tracking system to overcome existing field flaws and add full remote-control capability.

What I built

• Firmware for a distributed network of tracker controllers connected in a central server → Network Control Unit → tracker controller architecture.
• Integration of tilt sensors, motor drivers, motor-current sensing, RTC, and battery-voltage monitoring.
• XBee API-mode mesh communication with custom telemetry and heartbeat frames.
• Remote commands for setpoint updates, delays, GPS coordinates, calibration, cleaning, hazard mode, night mode, and fault clearance.
• Priority-based operation modes, calibration workflows, and detailed fault reporting (stall, reverse motion, slow movement, communication gaps, and more).
• Setpoint calculation on the tracker side using the NREPL backtracking algorithm.

Outcome

• Field deployment with 48 tracker controllers on a single coordinator.
• Auto command cycle every 30 seconds with periodic heartbeat reporting.
• Fully remote-configurable: setpoints, delays, and modes can all be updated from the central GUI.

Platform: STM32L4 series microcontroller
Role: Firmware Developer
Status: 🟢 Deployed in an operational polyhouse

An end-to-end environmental and crop-monitoring system built on a single board that gathers data from multiple industrial sensors and transmits it wirelessly.

What I built

• Integration of industrial RS-485 sensors for environmental, soil, crop-solution, and weather measurements: temperature and humidity, soil temperature and moisture, EC, pH, photosynthetically active radiation, dissolved oxygen, and an integrated ultrasonic weather station.
• RTC-based timestamping and XBee XR868 wireless transmission using custom data frames per board type.
• Cyclic external flash logging so data survives when the master goes offline. Unacknowledged frames are stored and re-sent automatically once the link returns.
• Remote configuration of RTC, transmission intervals, and flash memory.
• DIP-switch based device addressing for multi-board deployments.

Outcome

• Live dashboard with data logging and Excel export for analysis.
• Grafana cloud visualisations for long-term environmental trends.
• Stable in production for roughly 1.5 years without hardware changes.

Platform: STM32F4 series microcontroller
Role: Firmware + Desktop App Developer
Status: ✅ Completed

A configurable industrial gateway that bridges four serial ports to Ethernet with real-time, bidirectional data flow.

What I built

• Firmware for a TTL-to-Ethernet module integrated with an STM32 host controller.
• Each port independently configurable as RS-232 or RS-485 based on a 4-pin DIP switch, with per-port serial parameters.
• Non-blocking communication using interrupts and DMA across all four ports.
• RTC timestamping for every transmitted frame.
• A desktop application for live device data, port/module configuration, RTC setup, and data logging.
• AT-command-based configuration for the Ethernet module.

Platform: STM32L4 series microcontroller
Role: Firmware + App Developer
Status: 🔄 Under deployment

An eight-channel monitoring system for tracking string output currents in a solar plant, with live visualisation and logging.

What I built

• 8-channel monitoring for high-current solar strings.
• External ADCs over a shared I²C bus for multi-channel measurement.
• Analogue signal-conditioning circuitry (op-amp unity-gain follower) for bipolar Hall-effect sensor output.
• External flash storage for frames, RTC timestamping, and custom hex frames over Ethernet.
• Red/green LEDs per channel and a buzzer for fault indication.
• A desktop application with live current visuals, periodic polling, and Excel report export.

Platform: STM32L4 series microcontroller
Role: Firmware + App Developer
Status: ✅ Completed

A bridge device that receives XBee wireless telemetry and forwards it over Ethernet to a central system.

What I built

• XBee XR868 (API mode) integration with a TTL-to-Ethernet module.
• RTC-based timestamped transmission using a custom frame format.
• External flash memory for buffering and logging.
• 4-pin DIP switch for device addressing.
• A monitoring application with AT-command Ethernet configuration, RTC setting, and data logging.

Platform: STM32L4 series microcontroller
Role: Firmware Developer
Status: ✅ Implemented at a data centre

An autonomous thermal-management system for a data centre rack.

What I built

• Integration of a digital temperature sensor.
• Threshold-based automatic control of three 24 V DC cooling fans using a single sensor.
• Periodic, well-formatted ASCII telemetry over RS-485 for live monitoring via a serial tool.

Tech: Arduino · Proteus · Arduino IDE · ThingSpeak

A compact IoT system for monitoring air quality in dense urban pockets, with real-time remote visualisation.

What I built

• Sensor integration to monitor temperature, humidity, air quality index, and major pollutants such as CO, CO₂, NH₃, NOx, and particulate matter.
• Data pipeline to ThingSpeak for cloud visualisation and remote access.
• A small-form-factor design intended for deployment in tight urban areas.

Tech: Raspberry Pi · Servo Motors · IR Sensor · Solenoid Lock · HDMI Display
🏆 Sponsored by Techlance Pvt. Ltd.

Retrofitting a manually operated sanitary waste disposal machine into a fully automated, secure, and self-contained unit.

What I built

• Servo-driven metal door operations and an IR sensor for presence detection.
• Solenoid lock for secure, access-controlled disposal.
• An integrated HDMI display used for advertising and on-device usage guidance.
• A water-scrubbing air purifier with a mesh filter stage for air treatment around the unit.
• Recommended future scope: ventilation, isolated electronics compartment, maintenance alarms, and remote monitoring.

Tech: ESP32-CAM · LM35 · Peltier Module · Edge Impulse · Blynk IoT · NodeMCU

A smart cold-chain prototype that identifies produce and adapts cooling automatically.

What I built

• Object detection using an ESP32-CAM with a model trained on the Edge Impulse platform.
• Temperature sensing with an LM35 and adaptive cooling via a Peltier module, tuned per detected fruit or vegetable.
• IR-sensor-triggered operation and real-time remote monitoring via the Blynk IoT platform.