1) Micro-mobility: Sidewalk detection on Edge:

* Developed an IoT-based data-gathering platform by integrating an IMU sensor, camera, and speedometer on an ARM-based SBC incorporating Embedded Linux (ARM Cortex-A72) using I2C and serial communication respectively to record data at a rate of 125 Hz. [Language: C]

* Generated and trained a Random Forest Classifier on the collected data to extract features and classify the target values between street or sidewalk. The final accuracy of the model was reached out to be 95 %. [Language: Python]

*Developed generic scripts to read, write and parse data to the model, to upload recorded data on google drive using Google’s Pydrive API (on crontab), shell script to start the code at boot, etc.

• ---------------------------------------------------------------------------------------------------------------------------------------------

2) Firmware Testing Bike Docks:

* Responsible for generating test scripts to test the firmware for bike-sharing and holding devices of a multinational micro-mobility space company. The devices incorporated software that supports NFC-A (ISO14443A) and NFC-B (ISO14443B) communication for lock/unlocking mechanism.

* The test scripts were used for detecting continuous NFC tag reads, alternate locking/unlocking of NFC-A / NFC-B tags and then generating analysis reports and plots using matplolib. [Language: Python]

• Designed the RF controller to train the Automated Guided Vehicle (AGV) by establishing communication using LoRa, Arduino and Raspberry pi.

• Developed the C code to interface RF with Arduino using SPI communication. The Pi can be considered as a master and used to send commands serially to the Arduino.

• Designed the GUI for the RF controller using Tkinter library in Python for a 7” LCD touch screen interfaced with the Pi using I2C.

• ---------------------------------------------------------------------------------------------------------------------------------------------

• Designed a computer vision guided system to automate the process of a spray dryer using OpenCV and canny edge detection on NVIDIA TX2 (Master).

• Developed device drivers for UART, I2C, and ADC to interface the actuators and sensors with LPC1769.

• Developed a webserver on NVIDIA TX2 to host a webpage to display all the sensor values and control functions of the system.

• Set up communications between the NVIDIA TX2, LPC1769, and camera using TCP/IP communication protocol. [VG100]

*AIHS:

• Designed a mini-project with the title of Automated Indoor Herb Station using Arduino Uno R3 and multiple sensors and actuators.

• Developed an embedded C code using Arduino IDE to integrate moisture detection sensor, light sensor, plant-growing lamp, and water pump to automatically water the potted plants without any human interaction.

• The application was running without any complications in real-time.