The assignment analyzes the current state of mobile augmented reality (AR), highlighting its potential but also the limitations in achieving fast, precise, and comprehensive usability. It emphasizes the need for top-down approaches that integrate smartphone sensing, cloud offloading, and linear optimization to build effective AR applications. The core contribution discussed is a novel location-free geometric representation of the environment. Success metrics include latency, accuracy of object identification, ease of use, and scalability within controlled environments. A converged system using these principles is proposed as valuable for deploying engineering buildings.