RBE 594 Robotics Capstone Final Presentations

Join us via zoom on Tuesday, April 23 at 7:00pm (ET) for RBE 594 capstone presentations:

Join URL: https://wpi.zoom.us/j/95649805900

Autonomous Snow Removal Robot

Christian Chang, Stephen Crawford, Kyle Gloss, Colton Layhue, Rene Verduzco

Snow removal for homeowners has historically been a manual, arduous task. Snowblowers and other powered devices have eased this process, but snow removal continues to be a tedious, time-consuming, and potentially hazardous activity. Due to these drawbacks, the continuation of autonomous vehicle research and development, and ever-more affordable hardware required for these vehicles, an autonomous snow removal robot is not only viable, but desirable. In this presentation, we discuss some of the main subsystems required to make this device become a reality. The first of these subsystems includes the specifications of mechanical and electrical components needed to allow a robot to physically be capable of clearing snow from a residential property. Following this will be the methods for sensor fusion, integration, and calibration. These two subsystems create a foundation for the robot to achieve its goals and operate autonomously. This autonomous operation is then governed by three critical subsystems. The first of these subsystems is localization where the robot can use the data gathered by its sensors to find its global position and create a map of the world. Using this map and other known parameters of its environment, the second subsystem, motion planning, creates a trajectory for the robot to navigate through the world. The combination of both localization and motion planning is then given to its motion controller which finally energizes its movements and brings the robot to life. All of these subsystems in conjunction form the base for a completely autonomous snow removal robot capable of clearing residential driveways and sidewalks.

Intuitively Controlled Autonomous System to Aid the Visually Impaired

Jordan Grotz, Tom Mulroy, Sean Tseng, Max Weissman, Max Wolfley

The disparity between the demand for guide dogs and their availability highlights an urgent need for alternative solutions in aiding the visually impaired. Moreover, the working lifespan of guide dogs and their substantial cost pose significant challenges to accessibility. To bridge this gap, our project introduces an innovative solution: an intuitively controlled autonomous system to serve as a guide for individuals with visual impairments. This robot includes vision systems to track the user and obstacles in the environment. For global navigation, the system uses Google Maps, while locally, an artificial potential field algorithm fused with velocity obstacles path planning calculates finer paths. Enhanced mobility is achieved through motion planning with Mecanum wheels. Multi-modal LLMs are implemented for real-time human-oriented communication. Our solution proposes a more accessible, affordable, and versatile tool for people with visual impairments to navigate the world with independence.

Warehouse Material Transportation with Autonomous Mobile Robots (AMRs)

Jason Martino, Johnathon Fones, Joy Mehta, Leo Puerto, Philip Ni, and Sean McCormick

In response to the pressing challenge of a skilled labor shortage impacting supply-chain efficiency, workers find themselves increasingly burdened with non-technical tasks, notably the manual transportation of merchandise within warehouses. This not only poses safety risks but also consumes valuable time that could be better utilized for more critical responsibilities. Our project addresses this issue by introducing a robotic system designed to automate material transportation within factory environments, thereby alleviating this burden on skilled labor. Our primary objective is to enhance worker productivity by reducing their walking distance within the factory by at least fifty percent. Leveraging a model predictive controller and employing a differential drive system, our robotic solution optimizes path planning and navigation, ensuring efficient and safe movement of materials. By deploying algorithms, such as the A* global planner and the YOLOv8 feature detector for local object avoidance, our system optimizes path planning and navigation, ensuring efficient and safe movement of materials. Through a simulated environment, our project aims to illustrate that every unit of distance traveled by the robotic system translates into saved footsteps for workers, enabling them to focus on higher-value tasks. The technologies implemented for this proposed autonomous mobile robot demonstrates a potential system for the delivery of products throughout a warehouse to relieve tedious transport tasks from factory workers.