I am an Embedded Systems Engineer in Arizona
As a graduate student at Arizona State University majoring in Robotics and Autonomous Systems, I have gained valuable experience in designing, manufacturing, and programming through my various job positions.
Languages: Python, C++, MATLAB & Simulink, Embedded C, SQL, PowerShell
Software: Docker, ROS2, Solidworks, Autodesk Fusion 360, Arduino IDE, Altium, Microsoft Office
Hardware: Semtech SX12xx, NRF BLE, ESP32, SAMD21, Arm Cortex-M microcontrollers, ATmega, Raspberry Pi
Technologies: FreeRTOS, Gazebo, React Native, MQTT, Ubuntu, Tensorflow, Scikit-Learn, PyTorch
Protocols: SPI, I2C, CAN Bus, UART, RF integration (ZigBee, LoRa, Wi-Fi, BLE)
AWS: IoT Core, Lambda, Timestream, DynamoDB, S3 Buckets
• Engineered a self-sustained proprietary UHF off-grid full mesh protocol for ASU cart tracking and smart campus IoT initiative.
• Developed a Bluetooth Low Energy mesh network using ESP32s for SOS signals via a self-developed React-Native application.
• Wrote Lambda Functions with API Gateways and Timestream to obtain real-time location data as well as information about active
nodes around campus.
• Programmed mpu9250 for deep sleep acceleration-based interrupt and achieved 3 years of battery life for the cart tracker.
• Designed an articulated 3-axis linear cartesian robot attached to a 6-axis load cell with a closed-loop controller to perform tests that were used to determine the frictional characteristics of the gripper pads fabricated for the Lizard Inspired Tube Inspection (LTI) robot.
• Designed and fabricated gripper pads with curved textured surfaces using a Polydimethylsiloxane (PDMS) polymer to enable the LTI robot to perform friction-based mobility on curved surfaces irrespective of the material and surface texture.
• Headed and Co-founded the electric ATV team powered by a 8kWh BLDC Motor and a custom 48V Li-ion Battery pack.
• Incorporated 15+ sensors based on I2C and SPI communication protocol to collect data in real time.
• Simulated the vehicle’s performance on MATLAB and Simulink resulting in a 17% more efficient design.
• Compared various on-policy methods like DAPG, Monte-Carlo return methods like AWR to Advantage Weighted Actor Critic
giving 20% higher success rate.
• Reduced the time required to learn a range of robotic skills to practical time-scales by incorporating prior offline data along with
online tuning.
• Coded a Line Follower function for Parrot Mambo Mini-Drone.
• The drone used edge detection calculated the nearest edge by detecting specific HSV values of the track.
• The function was created on MATLAB and Simulink. It was then deployed to the drone via access point and bluetooth.
• To evaluate the performance of machine learning algorithms in detecting fraud in a bank transaction dataset.
• Investigated the impact of preprocessing technique, one-hot encoding with SMOTE, and perform feature importance analysis to identify the most significant features.
• Moreover, performed statistical tests to determine the significance of performance differences between the models and provide additional support for our findings.
• Designed a closed-loop PID controller for linear actuators that controlled the position of every individual linkage of a stewart platform to balance the motion of the ball placed on the platform. Reduced Steady State Error by changing the integral value.
• Developed a DAQ System to collect data from 12 sensors for data telemetry in real time using Arduino microcontroller.
• Integrated a GSM SIM 900 Module to Raspberry Pi Zero and transmitted sensor data using ThingSpeak Communication Library