Introduction
Exactly one week ago marked the start of my first summer research opportunity. The past seven days have served as a brief glimpse into the world of academia, and have set the scene for the next seven weeks of my life.
Research has always been something that fascinated me. The idea of being able to pursue knowledge for the sake of advancing our collective understanding of the universe as a career seems fantastical. However, after a few weeks of writing essays, applying to programs, and participating in interviews, I get to sample that dream at The Ohio State University.
I was recruited by Dr. Martijn IJtsma at OSU’s Cognitive Systems Engineering Laboratory to make contributions to the lab’s work in modeling and simulating work in human-machine teams (HMTs). Over the past few years, I’ve developed an affinity for computational modeling, especially modeling complex systems like work dynamics in HMTs. To say I was pleasantly surprised and excited when I initially received his email of expressed interest in my application would be a criminal understatement.
What the hell is Cognitive Systems Engineering?
… is the question that I initially asked myself when Dr. IJtsma reached out to me. I understood the description of his computational modeling work in his email, but the field of “systems engineering”, and especially “cognitive systems engineering,” was new to me.
The definition I was able to piece together after researching the laboratory and reading some of the related literature was as follows:
“
Cognitive systems engineering (CSE) is a discipline studying how people think and act in systems with many dynamic and interdependent parts (both human and automata). The goal is to design systems that best support people’s natural cognition in terms of decision making, problem solving, and task completion.
“
If you think that looks like a lot of jargon and overly academic terminology, welcome to the club. But to boil it down, we study how to make sure people can do their jobs in situations where they have to also work with other tools, people, and robots.
A good example, and Dr. IJtsma’s specialty, is aviation. Pilots have to coordinate with co-pilots, ground control, autopilot, and work with all of the other technology available to them. How can we minimize the cognitive demand of this system of making a plane get from A to B? This idea can be extrapolated to a ton of other important fields, like nuclear management (ever heard of Three Mile Island?), healthcare, or even outer-space operations.
So what do I do?
Dr. IJtsma and his team have been working with a simulation framework called Work Models that Compute (WMC). Essentially, it’s a method of simulating an environment where multiple workers (agents) have to complete individual tasks (take action), while working together. Agents are limited by the resources available in the environment, their dependencies on other agents, and their general capability in taking certain actions.
It’s an exceptionally useful tool for envisioning how to discover the necessary tools and agents for a job, and to determine who does what. As a very vague and general example, it can help determine if the work will be done faster if Agent A does Task 1 or if Agent B does Task 1.
My job is to work with Dr. IJtsma and a graduate student in the lab to use WMC to design a framework that can help determine the best sequence and timing of actions in a work environment. Agents are often interdependent, meaning they require other agents to finish their tasks or produce some resource before they can act. So if Agent A requires Agent B to do Task 1 before it can do Task 2, how can we schedule who does what and when to optimize performance?
It seemed a bit silly to me at first. “Just schedule it right after, obviously.” However, what if the agents have limited communication? What if there’s a third agent that also depends on Task 1, and needs the results sooner than Agent A does?
This computational analysis of task organization, agent interdependencies, and communication–coordination–is the heart of what I’ll be spending the next two months on.
Final Remarks
If my explanations weren’t entirely clear, I apologize. To be frank, I’m just starting to get a hold on the intricacies of the discipline and on the exact work I’m involved in. However, I understand a little more about CSE every day, and I hope to form a solid foundation in the coming weeks.
I am exceptionally excited and grateful to have this opportunity. I understand how lucky I am to be able to participate in professional academic research as a rising third-year undergraduate.
Familiarizing myself with the field’s literature and fundamental concepts over the past couple of weeks has been some of the most fun I’ve had reading and learning in years, and I’m confident my enthusiasm will only grow as I dive deeper into this project.
Thanks for reading!