Workshop @ CDC 2026

Learning-based and AI-enabled techniques are increasingly integrated into autonomous aerospace systems, bringing unprecedented performance capabilities together with new challenges in safety assurance, verification, and certification. Organized by the IEEE Technical Committee on Aerospace Control (TCAC), this workshop provides a focused forum for examining how formal methods and control-theoretic tools can support the trustworthy deployment of learning-enabled autonomy in safety-critical aerospace applications. Through invited contributions from leading researchers and industrial practitioners, the workshop highlights recent advances, open challenges, and emerging methodologies at the intersection of control, learning, and formal verification.

By combining academic perspectives with insights from industrial deployment and certification-oriented practice, the workshop aims to stimulate discussion on principled design workflows, assurance strategies, and future research directions for AI-in-the-loop aerospace control systems. The program is designed to foster interaction across communities and to encourage constructive dialogue on bridging theoretical developments and real-world implementation constraints.

The goal of this one-day workshop is to provide a comprehensive overview of current trends in the design of control architectures for autonomous aerospace systems, with a strong emphasis on formal methods, safety assurance, and industrial deployment. The program addresses both model-based and learning-based control strategies, with particular focus on verification, validation, and certification challenges arising in safety-critical, real-world applications. A central objective is to discuss representative industrial case studies and deployment scenarios, critically analyze the strengths and limitations of existing approaches, and highlight how formally grounded methods can support the reliable integration of learning-based components.

Recent developments at the intersection of control theory, learning methods, formal methods, and artificial intelligence have brought renewed attention to issues such as correctness guarantees, robustness, and certifiability of autonomous systems. These challenges are of paramount importance to the industry, where regulatory constraints and operational reliability play a key role in technology adoption. The strong participation of both academic and industrial speakers in this workshop reflects this growing convergence. For these reasons, we believe that the IEEE Conference on Decision and Control provides an ideal forum for bringing this timely and impactful topic to the attention of the control community and fostering dialogue among theory, tools, and industrial practice.

The full-day structure, extended discussion periods, and final round-table session are intentionally chosen to lower barriers to engagement, stimulate dialogue, and promote mentoring and knowledge exchange across career stages. In particular, exposure to industrial perspectives on certification and deployment offers valuable insight for young researchers and students who may otherwise have limited access to these discussions. In line with CSS’s commitment to broadening access and opportunity, the workshop fosters cross-disciplinary and cross‑sector interaction by bringing together communities spanning control theory, machine learning, formal verification, and aerospace engineering. This convergence supports the dissemination of best practices across fields and promotes transferable methodologies that can be adopted in adjacent domains beyond aerospace, including robotics, autonomous ground vehicles, and cyber‑physical systems. The organizers intend to make workshop materials (e.g., slides and summaries) broadly accessible after the event, facilitating wider dissemination of knowledge to participants unable to attend in person.

Unlike existing workshops that focus primarily on learning algorithms or autonomous applications, this workshop emphasizes certifiable, safety‑critical learning-enabled control architectures, integrating robustness theory, formal safety guarantees, industrial deployment constraints, and governance considerations within a unified framework. By addressing safety, certification, and governance of learning-enabled aerospace systems, the workshop contributes to the responsible adoption of autonomy in civil aviation and space systems, supporting public trust, regulatory acceptance, and the CSS mission of inclusive excellence, societal benefit, and responsible innovation.

This workshop will be sponsored by the IEEE Technical Committee on Aerospace Controls (TCAC). The invited speakers are mostly leading TCAC members from academia, industry, and government, actively involved in cutting-edge research and development in aerospace autonomy and control.

The tentative workshop schedule is reported in the Program page, together with the list of speakers and abstracts of the proposed contributions. Each talk will consist of 30' plus 5' for Q&A, separated by coffee and lunch breaks designed to maximize the participants' opportunities to discuss technical problems and tighten or establish professional networks.

The workshop is intended for Ph.D. students, researchers, engineers, companies, and practitioners who are interested in learning-based methods in aerospace systems and applications. With this initiative, the organizers aspire to establish an inclusive platform for engaging in open discussions with key stakeholders. Through its emphasis on safety, certification, and responsible AI deployment, the workshop aligns technical advances with societal needs, while its inclusive structure and diverse participation actively support the CSS mission of cultivating a vibrant, accessible, and forward-looking control community.

Expected outcomes of the workshop include a clearer understanding of how robustness and safety guarantees can be integrated with learning-based control architectures, increased awareness of certification and deployment constraints from an industrial perspective, and the identification of key open research challenges at the intersection of learning, control, and governance for safety-critical aerospace systems.