Robot Courier

Robot powered by 5 AA batteries
Ultrasonic sensor located in front of the robot

Inspiration Our journey began with the vision of creating a smart, autonomous delivery system that could navigate through dynamic environments with minimal human intervention. We were inspired by the growing need for automated solutions in logistics and personal assistance, and we set out to build a robot that could not only transport items but also interact intelligently with its surroundings.

What it does Robot Courier is designed to autonomously navigate its environment using a robust four-wheel drive system, controlled by a Raspberry Pi. The robot features an ultrasonic sensor for obstacle avoidance, ensuring smooth and safe navigation even in cluttered or unpredictable spaces. Additionally, we integrated a state-of-the-art face recognition module using PyTorch and MTCNN to identify users and personalize interactions. Although an unexpected camera failure prevented us from showcasing the face recognition feature, the system’s core functionalities remain fully operational.

How we built it Our build process combined both hardware and software innovations:

Hardware: We utilized a Raspberry Pi to serve as the brain of the robot, controlling a four-wheel drive chassis designed for stability and maneuverability. An ultrasonic sensor was integrated to detect obstacles in real time, ensuring safe navigation. Software: The robot’s control system was programmed using Python, where sensor data is processed to make real-time navigation decisions. For the face recognition component, we developed a module using PyTorch combined with the MTCNN framework to detect and recognize human faces. Despite the setback with our camera hardware, our development process provided valuable insights into machine learning integration in robotics. Integration: We focused on seamless integration of the mechanical and digital systems, enabling the robot to perform tasks autonomously and react to environmental changes with precision. Challenges we ran into Building Robot Courier was not without its hurdles:

Hardware Integration: Combining multiple sensors and ensuring reliable communication between components proved challenging. Fine-tuning the obstacle avoidance system required multiple iterations. Unexpected Setback: A significant challenge arose when our camera, a key component for the face recognition module, was accidentally short-circuited and burned out. This incident prevented us from demonstrating the face recognition functionality, despite our software being ready. System Coordination: Ensuring that the movement, obstacle detection, and (planned) face recognition features operated cohesively required extensive testing and calibration. Accomplishments that we're proud of We are particularly proud of:

Successfully developing a robust, four-wheel drive autonomous robot controlled by a Raspberry Pi. Implementing an efficient ultrasonic obstacle avoidance system that significantly enhances navigation safety. Designing and partially integrating an advanced face recognition module using PyTorch and MTCNN, showcasing our ability to merge robotics with machine learning. Overcoming multiple technical challenges, which enriched our understanding of both hardware and software integration in complex systems. What we learned Throughout this project, we gained invaluable experience in:

Embedded Systems: Mastering the intricacies of Raspberry Pi and sensor integration. Machine Learning Integration: Applying deep learning techniques for face recognition and understanding the challenges of deploying ML models on embedded hardware. Problem-Solving: Learning to troubleshoot and iterate on design flaws, especially when faced with unforeseen hardware failures. Team Collaboration: The importance of clear communication and coordinated effort in overcoming technical obstacles and achieving project milestones. What's next for Robot Courier Looking ahead, we plan to:

Replace and Enhance the Camera System: Address the hardware setback by sourcing a more robust camera module to fully realize the face recognition feature. Expand Functionality: Integrate additional sensors and features such as GPS for outdoor navigation and improved communication interfaces for real-time updates. Optimize Navigation: Refine the obstacle avoidance algorithms and incorporate machine learning techniques to adapt to varying environments. User Interaction: Enhance the interface for better human-robot interaction, making Robot Courier a more versatile and user-friendly assistant.