Conejo, Carlos (UPC), Alanwar, Amr (KTH), Puig, Vicenç (UPC), Morcego, Bernardo (Universitat Politècnica de Catalunya)

Integrating Signal Temporal Logic into Zonotopic Predictive Control for Autonomous Vehicles

Scheduled for presentation during the Invited Session "S17a-Synthetic-Data-Aided Safety-Critical Scenario Understanding in ITS" (TH-LM-T17), Thursday, November 20, 2025, 11:10−11:30, Southport 2

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 18, 2025

Abstract

Ensuring safety in autonomous driving over time requires control strategies that consider temporal requirements under uncertainty. This work presents a novel control framework that integrates signal temporal logic (STL) constraints into a zonotopic predictive control (ZPC). By embedding STL specifications into the data-driven predictive control formulation, the proposed method ensures that system trajectories satisfy not only dynamic constraints but also desired temporal properties, such as safety conditions and sequences of actions. Zonotopes are used to efficiently represent reachable sets under bounded uncertainties. The approach translates STL formulas into interval constraints over a finite prediction horizon, guiding the ZPC to meet the specified temporal requirements. The framework is validated using data from a real-world Renault Mégane autonomous vehicle (with SAE Level 3 of automation capabilities), proving its effectiveness in executing maneuvers that comply with formal temporal specifications.