Dr Jonathan Mayo-Maldonado

I am originally from Mexico, where I began my engineering training with a BSc and MEng in Electronic and Electrical Engineering at the Tecnológico Nacional de México. Those early years shaped habits that still guide my work: I state assumptions clearly, I model with care, and I let evidence decide. They also led me towards stressed operating regimes—situations where uncertainty grows, small neglected effects become important, and the difference between a clean diagram and a real device becomes visible.

In 2015, I completed my PhD in Electronic and Electrical Engineering at the University of Southampton (UK). My thesis, Switched Linear Differential Systems, received the Institution of Engineering and Technology (IET) Best UK PhD Thesis Award in Control and Automation. I am grateful for that recognition, but I value most what the process taught me: a result is only convincing when the chain of reasoning is clear and the limits are stated openly. In control, confidence is not enough; the argument must be checkable.

That view has shaped my research direction. I work at the intersection of control theory, power electronics, and power systems, with data-driven control at the centre. Classical control uses dynamical models—differential equations, state-space representations, poles and modes—to reason about feedback, stability, and performance. This language is powerful and still essential. However, modern energy systems often break the quiet assumption behind it: that a model can remain accurate as conditions change. Real plants heat up, age, drift, and operate across wide ranges; networks reconfigure; sensors degrade; disturbances do not arrive in neat patterns. In many practical settings, the model is not wrong, but incomplete.

Data-driven control starts from this practical fact. If the system changes, then measurements are not just for logging; they are evidence about how the system behaves now. This is not an excuse to abandon rigour. In fact, it forces more rigour, because data is imperfect. Measurements are noisy; sensors can fail; some operating regimes are missing; and rare events—the ones that decide safety—may not appear in the dataset at all. This leads to a precise question that guides much of my work: what can we guarantee from data, and what cannot we guarantee?

I treat that question as a technical programme. If we design control from measurements, we still have to answer the classic control questions, but we must answer them with explicit links to the data. Where is the “state” if we do not begin with a state-space model? What replaces poles and modes when we work directly with trajectories? When is a method truly data-driven, and when is it simply system identification followed by standard design? How much noise and sensor error can be tolerated before stability guarantees fail? I focus on methods that give clear answers, not only good numerical performance.

Energy systems make these questions urgent. Electrification and renewable integration are changing how power networks behave. As systems become increasingly inverter-dominated, stability and protection depend less on the passive dynamics of synchronous machines and more on control algorithms inside converters. In practice, this means that voltage and frequency behaviour is shaped by many controllers acting at fast time scales through sensing, computation, and communication. Stability is no longer only a property of the electrical network; it is also a property of software interacting with physics.

At the same time, the grid is becoming highly instrumented. Data streams from converters, machines, and networks can reveal couplings, regime changes, and early signs of instability that are hard to capture analytically. But large datasets can also create false confidence. They often describe normal operation well while under-representing stressed conditions, because operators avoid risky regions and because major failures are rare. For this reason, I work on approaches that combine model-based structure with data-informed design in ways that remain explainable and testable. Models help us state mechanisms and formulate hypotheses. Data helps us test those hypotheses, detect when behaviour has shifted, and quantify uncertainty using real operating records. The aim is control and monitoring that remain dependable when conditions change, not only when assumptions hold.

Once control depends on data streams, resilience becomes part of the control problem. Modern sensing, time synchronisation, communication, and software improve observability and flexibility, but they also create new dependencies. In inverter-dominated networks, a cyber incident can affect closed-loop dynamics by delaying measurements, corrupting signals, or manipulating setpoints. A controller that assumes clean data can be fragile. This is why I treat cybersecurity and resilience as integral to data-driven control and monitoring: credible methods must include credible failure analysis, detection mechanisms, and safe responses.

My work spans transmission and distribution networks, power electronic converters, electrical machines and drives, and energy storage systems. On the device side, I work on modelling, control, and diagnosis of machines and drives, and on control-oriented design and operation of converter systems. These topics connect directly to network-level dynamics because device behaviour often drives system-level phenomena. At the network level, I study modelling and analysis of converter-rich grids and develop monitoring and fault-detection methods for medium-voltage networks. Across these areas, I aim for methods that meet three standards: the mathematics is sound, the assumptions are explicit, and performance is tested under uncertainty, faults, and practical constraints—with a clear path to implementation and validation.

I have held academic posts at Tecnológico de Monterrey and the University of Sheffield. I have led and contributed to funded projects supported by SENER-CONACYT (Mexico) and UKRI, and I have published over 80 journal papers. Many contributions have come through collaborations with academic and industrial partners. I value this work because it keeps research grounded: deployment constraints reveal whether the bottleneck is sensing, computation, data quality, or modelling, and they show early where a method needs strengthening before it can be trusted.

Supervision and mentoring are central to my academic work. I have supervised and collaborated with more than ten PhD researchers, and I treat communication as part of research practice from the beginning. I try to build an environment where students can ask questions freely, present ideas clearly, and learn how to make claims that survive scrutiny. I enjoy working with researchers who care about mechanisms as much as outcomes, and who are willing to test ideas against critique, evidence, and implementation realities. A PhD, in my view, is training in judgement: knowing what can be claimed, what must be tested, and what should remain uncertain until the evidence is strong.

I am a Fellow of the Higher Education Academy (FHEA) and a Senior Member of the IEEE. I see these roles mainly as reminders of responsibility: to reason carefully, communicate precisely, and build work that holds up under stress.

Prospective PhD students interested in data-driven control, electrification, renewable integration, and cyber-secure control and monitoring of modern power systems are welcome to get in touch. If you want to work on methods that link measurements to stability and performance with clear assumptions and clear guarantees—and if you care about real deployment constraints as much as theory—we may be a strong match. The transition under way is technically demanding, and its failure modes are becoming more visible. Careful, transparent research can make a practical difference: not only by explaining what is changing, but by shaping solutions that are reliable, secure, and worthy of public trust.