Abstract
The way in which systems are controlled and regulated is an ever important issue in the world of engineering, and must constantly adapt in the face of technological developments. The former aspect of control is explored through my capstone group’s creation of a Rotary Inverted Pendulum (RIP), which makes use of control systems to actuate a free-spinning pendulum and respond to feedback in order to stand fully upright. The latter half about regulation is a main point of interest in my STS research paper on how artificial intelligence (AI) and machine learning (ML) is shaping the way vehicle manufacturers think about “safety” as this technology becomes more and more engrained into our daily lives.
My technical report explores the development of the RIP as an educational tool to support instruction in advanced control systems. Advanced concepts, such as control systems, can be difficult for students to learn purely through theoretical or mathematical approaches, so incorporating a physical model can enhance understanding by allowing students to visualize and apply these ideas in a real world context. The RIP is a mechatronic device that utilizes control systems– in this case, a Proportional-Integral-Derivative (PID) control system– to balance a pendulum rod in an upright, inverted position. Currently, the Department of Mechanical and Aerospace Engineering (MAE) does not have a RIP available for instructional use, motivating our capstone group’s efforts. This project and accompanying report detail the design and development of the RIP, including the iterative design process, technical challenges encountered, and analyses performed, culminating in a portable device for classroom use along with recommendations for future improvement.
My STS research paper examines how the concept of “safety” in modern vehicles is being reshaped by the rise of modern Advanced Driver Assistance Systems (ADAS) powered by stochastic AI/ML algorithms. While significant investment and technological progress have accelerated the development of AI/ML technologies in order to produce semi and fully autonomous driving systems, federal and state regulations have been struggling to keep pace, creating tension between perceived and actual safety. Through analysis of technical literature, regulatory frameworks, and consumer perspectives, this paper argues that safety is no longer a purely technical benchmark, but a dynamic construct influenced by engineering limitations, economic considerations, policy decisions, and societal acceptance.
