Using Ethereum for Secure Decentralized Optimization
Duration: 00:21:10
Speaker: Eric Munsing
Type: Talk
Expertise: Advanced
Event: Devcon
Date: Jul 2026
We demonstrate how complicated optimization problems can be solved by combining decentralized optimization algorithms with an aggregation step in a smart contract. Using tools from convex optimization, we decompose difficult problems into a set of subproblems with can be computed off-blockchain, finally reaching consensus on the global optimum by passing message with the on-blockchain aggregation step. We present an example of applying this approach to optimizing power dispatch on an electricity grid, but the approach can also be used to solve other problems in machine learning, coordinating robotic agents, or coordinating economic systems.