Enrique Mallada
Invited speaker · biography to follow
Details will be posted as the program is finalized.
Workshop overview
Modern dynamical systems increasingly face nonstationary dynamics, evolving objectives and constraints, heterogeneous uncertainties, and tight coupling with data-driven decision-making. Classical frameworks built on fixed models and static problem formulations are often insufficient. This workshop advances contextual control as a principled paradigm: policies explicitly depend on context, with emphasis on safety-critical domains such as power and energy systems, transportation, and robotics.
We bring together researchers from control theory, optimization, and learning to clarify foundations, surface theoretical challenges, and showcase emerging methods and applications in large-scale engineered systems.
Objectives
Establish contextual control as a shared language within the IFAC community, linking classical methods with learning and guarantees.
(i) Foundations
Define and formalize contextual control as an extension of classical control, optimization, and learning.
(ii) Theory
Identify key challenges: stability, robustness, generalization, and guarantees across contexts.
(iii) Applications
Highlight emerging use cases in safety-critical systems: power and energy, transportation, robotics.
(iv) Community
Stimulate directions that integrate learning with structure, physics, and control-theoretic guarantees.
Topics of interest
Foundational theory, methodology, and applications—including but not limited to:
- Foundations of contextual control. Formalizing “context”; relationships to gain scheduling, adaptive control, and hybrid systems; hierarchical and multi-timescale views.
- Context-aware optimization and control. Context-dependent MPC; robust and distributionally robust control under context-dependent uncertainty; switching and hybrid control with regime-dependent dynamics.
- Learning-based contextual control. Contextual and conditional reinforcement learning; meta-learning and transfer for control; learning with structure-preserving control laws.
- Context, uncertainty, and decision dependence. Decision-dependent uncertainty and endogenous randomness; uncertainty models conditioned on actions; risk-sensitive and reliability-aware contextual control.
- Applications in complex systems. Power and energy (renewables, DERs, markets, frequency/voltage control); transportation and autonomy; robotics and human–machine interaction; large-scale infrastructure and networks.
Tentative schedule
Sunday, August 23, 2026 — full-day program (times in local congress time). Talk titles and speakers to be announced.
09:00–09:40
Talk 1
09:40–10:20
Talk 2
10:20–10:40
Coffee break
10:40–11:20
Talk 3
11:20–12:00
Talk 4
12:00–14:00
Lunch break
14:00–14:40
Talk 5
14:40–15:20
Talk 6
15:20–15:40
Coffee break
15:40–16:20
Talk 7
16:20–17:00
Talk 8
Invited speakers
Internationally recognized researchers across control theory, optimization, learning, and applications. The program emphasizes complementary perspectives on foundations, methods, and open problems.
James Anderson
Invited speaker · biography to follow
Details will be posted as the program is finalized.
Richard Pates
Invited speaker · biography to follow
Details will be posted as the program is finalized.
Additional invited speakers may be listed in the final congress program. The proposal emphasizes diversity in gender, geography, institution, and career stage to support inclusive discussion within the global IFAC community.
Organizers
Audience & outcomes
Who should attend
- Researchers in control theory, systems and control, and optimization
- Learning-based control, RL, and data-driven decision-making with interest in guarantees
- Application researchers in power and energy, transportation, robotics, and large-scale CPS
- Graduate students and early-career researchers exploring integrated paradigms
Background
- Linear systems, feedback control, stability
- Optimization and optimal control
- Dynamical systems and modeling
- Machine learning and reinforcement learning
Expected outcomes
- Shared understanding of contextual control within IFAC
- Clearer links to classical control, optimization, and learning-based methods
- Identification of open problems and emerging applications
- Connections for future special sessions, journal issues, or follow-on workshops
Timeliness
Renewables-heavy grids, large-scale infrastructures, and autonomous systems challenge fixed-model paradigms, while ML/AI is often deployed without full integration of physics, stability, or constraints. This workshop targets that gap with a framework that unifies learning, optimization, and control theory—aligned with IFAC’s mission and classical themes such as adaptive, robust, hybrid, and optimization-based control.