Behavioral Analysis of Smart Contracts Using Supervisory Control Theory

Licentiate thesis, 2026

Smart contracts are programs deployed on blockchains that are typically used to manage distribution of digital assets and cryptocurrencies. Smart contracts are used in applications such as auctions, games, financial services and decentralized applications. It is not possible to change a smart contract after it has been deployed on the blockchain network. Since smart contracts can hold and manage transfer of digital assets, even a small error in their behavior can lead to serious consequences. This thesis explores how smart contracts can be modeled as event driven systems and be analyzed for certain properties. We present a framework for automatically converting Solidity smart contracts into Extended Finite State Machine models that describe the execution of the smart contract. Using these models, we show how non-blocking verification can detect situations where a contract may fail to reach its intended completion. In particular, we study cases where failed transfers may result in funds being locked or prevent the expected progress of the contract. Furthermore, this thesis investigates bias in smart contracts. Bias here refers to an advantage/disadvantage allowed by the behavior of the contract to a user, or a group of users. Such an advantage/disadvantage reveals the ability to force an outcome that either benefits the user(s) or harms others. By analyzing the contract from perspectives of different users, we show how such a bias can be detected.