About Shauna Heron

Data science · Experimental psychology · Human–robot interaction

A bit of the longer story

My favourite kind of problem is messy. Give me a giant health record dataset with 13 possible joins, a pile of wonky Excel files, a robot that only behaves on a strict set of rules when you need it to do more, or a question a clinical team has been asking for years — I want to figure out which code or tools will turn that mess into something actually useful.

I came to research the long way around: early-web work in the Netherlands, a stint as a youth officer in secure custody, a decade running an award-winning wedding-photography business. Each was its own version of the same passion — learning a new system and paying close attention to how people actually behaved inside it. That’s still what my research is about, whether the system is a machine-learning model running on clinical data or a small robot trying to hold a conversation.

What I’m working on right now

• Finishing my master’s thesis — a machine-learning analysis of clinician workload in community mental health care — and preparing for my oral defense.
• Raising a teenager in the wilds of northern Ontario, which is a full-time job in itself.

What I’m curious about

• Applications of data science in public and population health–especially mental health, and especially in community settings. I want to know how we can use data to make care more effective, efficient, and equitable, without adding to the burden on clinicians or patients.
• When is a machine-learning model the right tool for a clinical or operational problem, and — more often — when is it the wrong one? I’m interested in interpretable models, fairness audits, and being honest about what a model can’t do.
• How do we design robots that people can actually work with? I want to understand how people form trust in autonomous systems, and how we can design those systems to be trustworthy in the real world.