Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AI4OPT: AI Institute for Advances in Optimization

Pascal Van Hentenryck, Kevin Dalmeijer

TL;DR

AI4OPT presents a framework to fuse AI and optimization for scalable engineering decision-making, addressing challenges in supply chains, energy systems, chip design, and sustainability. It introduces optimization proxies with a feasibility repair layer and end-to-end learning, detailing seven methodology thrusts guided by four main end-use cases. The paper emphasizes education via a 'teach the teachers' model and robust workforce development, alongside industry partnerships to accelerate impact. Overall, the approach aims to deliver real-time risk assessment, resilient planning, and scalable, equitable AI-driven optimization across critical sectors.

Abstract

This article is a short introduction to AI4OPT, the NSF AI Institute for Advances in Optimization. AI4OPT fuses AI and Optimization, inspired by end-use cases in supply chains, energy systems, chip design and manufacturing, and sustainable food systems. AI4OPT also applies its "teaching the teachers" philosophy to provide longitudinal educational pathways in AI for engineering.

AI4OPT: AI Institute for Advances in Optimization

TL;DR

AI4OPT presents a framework to fuse AI and optimization for scalable engineering decision-making, addressing challenges in supply chains, energy systems, chip design, and sustainability. It introduces optimization proxies with a feasibility repair layer and end-to-end learning, detailing seven methodology thrusts guided by four main end-use cases. The paper emphasizes education via a 'teach the teachers' model and robust workforce development, alongside industry partnerships to accelerate impact. Overall, the approach aims to deliver real-time risk assessment, resilient planning, and scalable, equitable AI-driven optimization across critical sectors.

Abstract

This article is a short introduction to AI4OPT, the NSF AI Institute for Advances in Optimization. AI4OPT fuses AI and Optimization, inspired by end-use cases in supply chains, energy systems, chip design and manufacturing, and sustainable food systems. AI4OPT also applies its "teaching the teachers" philosophy to provide longitudinal educational pathways in AI for engineering.
Paper Structure (17 sections, 4 figures)

This paper contains 17 sections, 4 figures.

Table of Contents

  1. Introduction
  2. Optimization Proxies
  3. End-Use Cases
  4. Supply Chains
  5. Energy Systems
  6. Chip Design and Manufacturing
  7. Sustainable Systems
  8. Methodology Thrusts
  9. End-to-End Optimization
  10. New Generation Solvers
  11. Decision Making Under Uncertainty
  12. Reinforcement Learning
  13. Combinatorial Learning
  14. Distributed and Multi-Agent Learning and Optimization
  15. Ethical AI
  16. ...and 2 more sections

Figures (4)

  • Figure 1: The Architecture of Optimization Proxies.
  • Figure 2: Real-Time Risk Assessment.
  • Figure 3: Faculty Training Program Cohort 1.
  • Figure :