I’m an Assistant Professor of Electrical Engineering and Computer Science at the University of Michigan, working on distributed control theory with applications to neuroscience, biology, and engineering systems. I received a BASc in Engineering Science from the University of Toronto in 2018, and a PhD in Control + Dynamical Systems from Caltech in 2023. Outside of work, I enjoy playing, composing, producing, and listening to music. I’ve also had the joy of working as a private piano and vocal music instructor for children and adults.

[CV] [Google Scholar] [GitHub] [LinkedIn] [Miscellany]

Open access to research is important to me. I endeavor to make some version of all my papers available on arXiV or biorXiV — these should be easily found via Google Scholar.

My pronouns are she/her/hers.

For prospective students

I am currently open to hiring students. If you are interested in joining my group, you should have experience or coursework in control theory. You should also have some interest in neuroscience and biology — background is not required, but curiosity and willingness to learn are required! Apply to the ECE or CSE program and include my name somewhere (e.g. in your research statement) as an advisor you are interested in working with.

If you are currently a student at U of M and are interested in my research, I encourage you to take or audit my topics class, offered winter 2024.

I can be reached at jslisali at umich dot edu. Please read the below first if you plan to email me:

Emails I will respond to (typically within a week):

• Questions about research and planned projects. Please include a description of your own research interests, background, etc.
• Questions about what it’s like to work in my group. Some information on my mentorship style can be found here.

Emails I will generally not respond to:

• Questions about how to apply to the PhD program. Instructions can be found on the departmental webpage.
• Questions about the suitability or status of your PhD application package. These are handled on a departmental level, and you will be notified according to standard procedures.
• Requests to meet (in-person or virtual), unless you are already at U of M.
• Other low-effort correspondence, e.g. emailing me without any idea of my research areas, clear research mismatch, etc.

Research overview

In engineering, control theory plays a crucial role in the design and analysis of robust and efficient systems — including robots, spacecraft, and power grids. In neuroscience, control theory underlies models of motor and sensorimotor function. Control theory also has applications as a modeling tool for various biological processes, such as metabolic dynamics. I am interested in control-theoretic analysis of biological systems, which yields novel scientific explanations as well as valuable design insights, which can be applied to engineering systems.

The three foundational directions of my research are:

(1) develop theory for the design and implementation of controllers under various energy and signaling constraints

(2) apply theory to develop novel models and insights for living systems, and

(3) apply theory to create scalable, robust, efficient algorithms and architectures.

I am particularly interested in how biology co-designs and implements controllers using organic components, which face unique challenges compared to their electronic counterparts.

The marriage of control theory and biology is far from new, but it is an unfamiliar union to many engineers and scientists. For engineers and control theorists who wonder why one might choose biology as a main application, I offer up the following explanation. Conversely, for neuroscientists and biologists who are curious about controls, I elaborate on some unique benefits of control theory as a modeling tool here. These snippets are adapted from my PhD thesis.

Distributed control

My theoretical research focuses on distributed, scalable control algorithms. These algorithms accommodate communication constraints, and enjoy complexity that scales independently of network size under mild assumptions. This is important for large-scale systems, for which centralized controllers can be intractable. I often use the system level synthesis framework in my research — I write and maintain toolboxes in MATLAB and Python for this framework. Representative works include scalable structured robust control, model predictive control (1, 2, 3), and closed-loop and controller specifications for distributed algorithms. These works may also be applied to models of neuroscience and biology, in which distributed operation and communication constraints play a key role.

Recent work in neuroscience

Internal feedback in sensorimotor systems

The standard model of sensorimotor processing involves signal flows from sensory inputs to sensory areas, then from sensory areas to motor areas. However, massive amounts of signal flow in the opposite direction (i.e. from motor areas back toward sensory areas) are observed in the cortex; we refer to these signals as internal feedback. What function does internal feedback serve, and why does the brain contain so much of it? We answer these questions using control theory. In this set of three papers geared toward control audiences, we outline the presence of internal feedback across a variety of organ systems and organisms, and apply both standard control theory and distributed control theory to analyze internal feedback. In particular, we argue that the incorporation of local communication and localized behavior necessitates large amounts of internal feedback. We also explore additional connections to existing physiological and experimental data in this paper geared toward a neuroscience/general audience. This work offers the first explanation for the large quantities of internal feedback observed in the cortex. Local and distributed processing also feature heavily in the analysis of chemical reaction networks, which is the subject of ongoing work.