STEM II is taught by Dr. Crowthers. Students in this course design an assistive technology device for a client whom they work closely with and receive feedback from. Students must utilize the engineering process to design, test, and modify prototypes until landing on a final, deliverable product.
SEARCH: Smart Electronic Assistance and Retrieval Companion for Home
For our STEM II project, I served as the Chief Executive Officer (CEO) and worked with Chief Informational Officer (CIO) Charles Tang, Chief Technical Officer (CTO) Tarun Eswar, and Chief Manufacturing Officer (CMO) Nevin Thinagar.
In the United States, over 50% of the elderly population struggle with mental impairments such as dementia, memory decline, and conditions of forgetfulness. Mental impairments and memory decline in the elderly can have significant impacts on their daily lives. Tasks that were once simple and routine can become challenging and frustrating, leading to feelings of helplessness and dependence on others. These mental impairments cause not only a decrease in the quality of life in the elderly, but often cause embarrassment as they cope with growing forgetfulness. Furthermore, as the age of a person increases, their short-term memory capabilities decline, such as in the ability to remember where they placed an important item.
Our design will be a rover-based device equipped with computer vision technologies that can be a powerful tool for locating misplaced items. The device can navigate autonomously through an environment on the ground while using its onboard cameras and advanced object recognition algorithms to identify and locate specific objects. Once an item has been located, the device can alert the user to its location through an application. This technology has the potential to save time and reduce frustration by quickly locating lost or misplaced items.
Our design is an autonomous rover device equipped with computer vision technology to efficiently locate lost or misplaced items. It navigates through various environments using onboard cameras and advanced object recognition algorithms, providing users with quick and accurate identification of specific objects. The rover chassis, developed in collaboration with a 4WD Smart Car Chassis for Raspberry Pi, features rubber wheels with sponge inserts for improved suspension and traction on slippery surfaces. The Raspberry Pi Model 3B+ is the central computing hub, securely attached to the PVC chassis and configured for wireless communication. The 3D-printed camera mount, assembled with screws and super-glue, allows easy attachment to the rover body. Object detection is achieved through a state-of-the-art image processing model that recognizes 1000 object classes deployed using Tensorflow Lite and PyTorch. A master script controls the model, enabling users to test its capabilities.