Learning to solve complex tasks with signal temporal logic (STL) specifications is crucial to many real-world applications. However, most previous works only consider fixed or parametrized STL specifications due to the lack of a diverse STL dataset and encoders to effectively extract temporal logic information for downstream tasks. In this paper, we propose TeLoGraF, Temporal Logic Graph-encoded Flow, which utilizes Graph Neural Networks (GNN) encoder and flow-matching to learn solutions for general STL specifications. We identify four commonly used STL templates and collect a total of 200K specifications with paired demonstrations. We conduct extensive experiments in five simulation environments ranging from simple dynamical models in the 2D space to high-dimensional 7DoF Franka Panda robot arm and Ant quadruped navigation. Results show that our method outperforms other baselines in the STL satisfaction rate. Compared to classical STL planning algorithms, our approach is 10-100X faster in inference and can work on any system dynamics. Besides, we show our graph-encoding method's capability to solve complex STLs and robustness to out-distribution STL specifications. Code is available at https://github.com/mengyuest/TeLoGraF

Many robot tasks have complex instructions that include both timing and the goals, like "reach the red zone within 5 seconds, but avoid obstacles at all times". These instructions can be written in a formal language called Signal Temporal Logic (STL). But current AI models can only handle simple or fixed STLs. We developed TeLoGraF, a novel approach to learn solutions for general STL forms. We created a large dataset of 200K STL instructions paired with example solutions, and used a graph-based neural network to help the AI reason about the logic and generate the solutions.