Zaid Memon Email & Phone Number

Karachi, PK

Karāchi, Sindh, Pakistan

• Directed a cross-functional team of 15 members and 4 departments, over a 9-month project cycle, to design and implement autonomous system architecture for Pakistan’s first ever driverless vehicle while managing a.
• Planning and developing strategy, handling external and internal affairs, and overlooking finances of the team, responsible for securing an international sponsorship worth PKR 1.8 million.
• Maintained Git workflows for… Show more
• Maintained Git workflows for multiple software projects, resulting in streamlined collaboration and seamless code integration among team members. Show less

• Learned ROS (Robot Operating System)
• Basic concepts of robot kinematics and dynamics
• 3D pose estimation using a neural network-based approach for monocular vision
• 3D pose estimation using a clustering-based approach for stereo vision
• Implementation of a custom RRT algorithm for path planning
• Implementation and Tuning of PID, Stanley and Pure Pursuit controllers

• Developing and managing the team's website
• Designing posters and creating video content for the team's social media platforms
• Planning outreach campaigns
• Conducting marketing calls and assisting with business presentations
• Understanding social media algorithms to develop strategies for enhancing the team's presence

The project aimed to implement an efficient and robust 3D SLAM based approach on an omni-directional wheeled mobile robotic platform in order to enable true autonomous navigation in an unpredictable and previously unknown environment where GPS data is not available, such as indoors.

Wooler, England, United Kingdom

• Implemented a reliable solution for detecting water bodies in off-road environments, achieving an F1 score of 0.85, through the utilization of an SVM-based system that leveraged color and textual features of water.
• Employed a preprocessing routine to enhance image quality by 15% and boost the accuracy of a segmentation model, through the implementation of automatic exposure adjustment for a moving camera.

• Led the design and implementation of an autonomous and decentralized search and rescue mission using a swarm of multi-rotor UAVs, achieving 18th position among 98 teams at TeknoFest 2022.
• Improved inference time of object detection and classification models by 4x through the deployment models using TensorRT on Jetson Nano, contributing to NUST Airworks' 2nd place finish in the 2021 IMechE UAS Challenge.
• Designed an algorithm for efficient search area coverage using optimum… Show more
• Designed an algorithm for efficient search area coverage using optimum trajectories for drones, with a GUI for easy mission planning. Algorithm was able to adapt to varying search area sizes and shapes. Show less

• Image processing using NumPy and OpenCV, utilizing various functionalities of the libraries including the GPU and DNN modules
• Alphanumeric character classification leveraging transfer learning using VGG-16, VGG-19, MobileNetV1, ResNet50V2, and EfficientNetB0
• Object detection using the YOLOv3, and Scaled YOLOv4 (P5)

• Cofounded the society and built its software department from the ground up, effectively recruiting and training 10 members to form a cohesive and skilled team.
• Developed the software for a social robot, "Cy", featuring advanced conversational AI, speech recognition, and speech synthesis capabilities with a sound source localization success rate of 95% for 120-degree head.
• Programmed robots for line-following, and firefighting, resulting in a 2nd place win in the… Show more
• Programmed robots for line-following, and firefighting, resulting in a 2nd place win in the firefighting category at RoboFiesta 4.0 competition Show less