 This paper proposes a new approach to controlling assistive devices for individuals with tetraplegia. It uses shoulder movements to detect user intent and then translates them into commands for the device. This allows for more intuitive and natural control of the device, which is important for those who have limited or no-hand function. The system was tested with 10 able-bodied and two tetraplegic subjects, achieving an average classification accuracy of 80% and 84%, respectively. These results show that the system is reliable and could potentially provide a viable alternative to existing methods of controlling assistive devices. This article was authored by Lucas Fonseca, David Garaud, Arthur Hierassery, and others.