In our lab, we aim to address the challenges in the following research thrusts:
1) Resilient Autonomy: Failures are inevitable in real-world applications. Autonomous systems integrated in our daily activities should not behave randomly in the case of such failures. One potential way of avoiding this is enforcing an idle mode or a pre-defined behavior to the system while the specification is revised and the violation is resolved. However, the duration of revising a specification (e.g., from operator feedback) or resolving the violation can be arbitrarily long. In this research thrust, we aim to address the following research question: how can we allow a system to continue execution with guarantees in terms of the satisfaction of relaxed specifications (e.g., drone lands at a site close to the desired end location in case it is occupied, autonomous car continues driving to find another opportunity to change lanes when the lane is not changed in the desired time)? For example, we investigated resilient operation of autonomous vehicles for urban air mobility concept, and we developed a priority-based algorithm, which guarantees safety, resolves any deadlocks, and ensures the completion of a spatio-temporal task with a finite delay (supported by Honeywell Aerospace and MnDRIVE).
Some related papers:
- R. Peterson, A. T. Buyukkocak, D. Aksaray, and Y. Yazicioglu, “Safe Reactive Motion Planning Using Time Window Temporal Logic Specifications”, Robotics and Autonomous Systems, vol. 142, pages 103801, 2021.
- D. Aksaray, "Resilient Satisfaction of Persistent and Safety Specifications by Autonomous Systems", AIAA Scitech Forum, 2021.
- D. Aksaray, C.I. Vasile, and C. Belta, “Dynamic Routing for Energy-Aware Vehicles with Temporal Logic Constraints”, IEEE International Conference on Robotics and Automation, Stockholm, Sweden, 2016.
2) Constrained Reinforcement Learning: Constraint satisfaction is an important aspect in the process of learning optimal policies. When an autonomous vehicle is simultaneously learning and performing its mission in real time, it is not acceptable to violate constraints during the exploration process. Exploration with possible constraint violation might potentially cause catastrophic events such as accidents or undesired performance. In this research thrust, we aim to develop learning algorithms that yield optimal policies while ensuring not to violate desired spatio-temporal specifications throughout learning. This research is currently supported by DARPA DSO.
Some related papers:
- D. Aksaray, Y. Yazicioglu, and A.S. Asarkaya, “Probabilistically Guaranteed Satisfaction of Temporal Logic Constraints During Reinforcement Learning”, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021.
- A.S. Asarkaya, D. Aksaray, and Y. Yazicioglu, “Temporal Logic Constrained Hybrid Reinforcement Learning to Perform Optimal Aerial Monitoring with Delivery Drones”, International Conference on Unmanned Aircraft Systems (ICUAS), 2021.
3) UAV-UGV Teaming: One of the fundamental research questions in the area of multi-robot systems is how to achieve global objectives with a group of heterogenous agents, each of which has different capabilities and act based on some partial information about the overall system (limited sensing/communication/computation). In this research thrust, we aim to address the research challenges in information-gathering problems via multi-agent aerial robotic systems (e.g., aerial monitoring with drones) both from a theoretical and experimental perspectives.
For example, we proposed a hierarchical architecture where a high-level swarm objective is decomposed into individual temporal logic formulae such that each agent in the swarm aims to synthesize a controller that satisfies its temporal logic (supported by DARPA Offset program). We are currently investigating the coordination of heterogenous multi-agent systems for space exploration missions (currently supported by NASA JPL). Moreover, we are developing planning algorithms that optimize the trajectories of drones and mobile charging stations simultaneously while ensuring safety (e.g., drones never run out of energy during flight, obstacles/collisions are avoided).
Some related papers:
- A.T. Buyukkocak, D. Aksaray, and Y. Yazicioglu, "Planning of Heterogeneous Multi-Agent Systems under Signal Temporal Logic Specifications with Integral Predicates", IEEE Robotics and Automation Letters (RA-L), vol 6, no 2, 1375-1382, 2021.
- Y. Yazicioglu, R. Bhat, and D. Aksaray, “Distributed Path Planning for Serving Cooperative Tasks with Time Windows: A Game Theoretic Approach”, Journal of Intelligent & Robotics Systems, vol. 103,(2), 1-19, 2021.
- A.T. Buyukkocak, D. Aksaray, and Y. Yazicioglu, "Distributed Planning of Multi-Agent Systems with Coupled Temporal Logic Specifications", AIAA Scitech Forum, 2021.
- X. Lin, Y. Yazicioglu, and D. Aksaray, "Robust Planning for Persistent Surveillance with Energy-Constrained UAVs and Mobile Charging Stations", IEEE Robotics and Automation Letters (RA-L), 2022.
- S. Seyedi, Y. Yazicioglu, and D. Aksaray, “Persistent Surveillance with Energy-Constrained UAVs and Mobile Charging Stations”, IFAC Workshop in Distributed Estimation and Control in Networked Systems, Chicago, IL, 2019.