250 likes | 324 Views
OwlSim is a cutting-edge simulation framework developed in collaboration with Citizens for Affordable Energy to model and evaluate long-term energy policies. The platform aims to provide accessible results to the public and allow experts to modify assumptions and models. The theoretical design involves modular circuit elements that can be composed to create complex models. The simulation framework supports multiple users, scales seamlessly, and offers various use cases for modeling and analysis. OwlSim enables the comparison of high-level energy plans and generates worst-case, average-case, and best-case scenarios for informed decision-making. The conclusion includes acknowledgments for CFAE, Microsoft, and individuals who contributed to the project. Potential questions are addressed, such as the choice of Azure cloud, development challenges, and the utilization of historical data for model development.
E N D
OwlSim: Revolutionizing National Energy Policies Through Technology COMP 410 in Collaboration with Citizens for Affordable Energy
Overview • Introduction • COMP 410 • Project Motivation • Project Objectives • Simulation Framework • Energy Model and Plans • Demo • Conclusion • Questions
About COMP 410 • “Software Engineering Methodology” • Design class satisfying computer science Bachelors of Science Decree capstone requirement • Typically taken by juniors and seniors • Warm-up project during first 3 weeks, then semester-long project … with a real customer! • Student driven – no problem sets or lectures
Our Customer:Citizens for Affordable Energy • CFAE is a national not-for-profit membership association • Goal is to educate citizens and policymakers about non-partisan national energy solutions • Leadership • John Hofmeister, Founder and CEO • Karen Hofmeister, Executive Director
Project Motivation • U.S. has no long-term national energy policy • CFAE believes this could be disastrous • CFAE wants a software tool that can simulate the long term effects of policies (or lack thereof) • Shifting policy landscape mandates a very flexible simulator
Our Mission • Create a mathematical model of the U.S. electric power generation and distribution • Make a set of policies and assumptions for the next 50 years of the energy industry • Develop a simulation framework to evaluate the effects of the plans based on the assumptions and the model • Make the results accessible to the public • Make the framework flexible and extensible to allow domain experts to change our assumptions and models
Theoretical Design • Modeling complex systems with mathematical functions • Functions represented as modular “circuit elements” with inputs and outputs • Functional modules can be “composited” • Encapsulate components of model • Allows composite modules with other modules inside. • Arbitrarily complicated models can be created • Diagram!
Capabilities • Supports many simultaneous users • Scales with load • Basic use case • View model, plan, precomputed results • Authenticated use case • Edit plan, recompute results, save results • Expert Authenticated use case (if working) • System Administration use case (if working) • Publish results (if working)
Demo • Connecting through web • Explain GUI • Basic use case • View model, plan, precomputed results • Authenticated use case • Edit plan, recompute results, save results • Expert Authenticated use case (if working) • System Administration use case (if working) • Publish results (if working)
Acknowledgements • Citizens for Affordable Energy, John Hofmeister, Karen Hofmeister • People we consulted for help • Microsoft for sponsorship • Steven Wong, Scott Rixner • TAs: • Team leads and all our hard-working students
Questions • Can’t this be done in Excel? • Yes, simple test model could be done in Excel, but professionals could develop more complicated models. Excel doesn’t have centralized mechanism for managing results. Visualizing model, others can edit, maintainability, portability. Results is not only objective – also export. • How can you develop model without domain knowledge? • Talked to experts, used recognized resources (EIA etc.). Simple but reasonable model. • How did you choose Azure over other clouds? • Had past support from Microsoft
Questions • What was biggest technical challenge? • What was biggest non-technical challenge? • HTML5, other technology choices • Did you use curve fitting on historical data?
References • EIA etc.