Enhancing Robotic Software Through Quality Intelligence: Data-Driven Approach to Optimization & Performance Improvement

Robotic software plays a critical role controlling hardware and enabling robots to interact with the physical world to perform complex tasks. To ensure reliability, the software must evolve to address sensor errors, hardware malfunctions, user errors, and environmental uncertainties. This paper explores how the integration of Quality Intelligence (QI) offers a structured, data-driven approach to improving robot performance, reliability, and real-time decision-making. Drawing from the experience of the FTC team Revamped Robotics, this paper outlines key challenges and practical solutions in enhancing robot software.

FTC is a global robotics competition for pre-college students, focused on real-world problem-solving. The team applied sensor-based data collection, automated testing and tuning, iterative design, and predictive modeling to systematically improve robot efficiency. Using encoders, IMUs, and cameras, along with data logging, the team tracked robot movement, speed, and measured errors. Vision processing enabled object detection and real-time adjustment. Specialized automated tests improved autonomous path-following, while experiments guided parameter tuning for consistency. Virtual physics-based simulations based on predictive modeling streamlined testing and design decisions.

By adopting these advanced methodologies, the team demonstrated how QI can serve as a catalyst for smarter, more efficient robotics development.

Paper | Presentation

Havish Sripada

Havish is a high school student at Jesuit high school and the lead programmer of 12808 Revamped Robotics, FTC team.

Sophia Lee


Jayden Mei


Thilan Wijeratne