I learned a ton from doing this. I figured out how to actually build one of these Human-Robot Interaction things, connecting the FurHat robot to a sma

A FurHat-based System for Interactive Interview Training: Comparing Virtual and Physical Human-Robot Interaction

I got into this project because I saw how tough the job market is and how much people stress about interviews. I thought it'd be cool to use the FurHat robot to make a new kind of interview practice tool, something better than what's out there. I'd seen AI interview platforms, but they felt a bit flat without a physical presence and that real-time back-and-forth you get with a person. So, I wanted to mix the smarts of AI with the physical side of robots to make the whole practice experience more real.

The project definitely had its tricky bits, what I called "Coding Traps" in the dissertation. One thing was the OpenAI API being slow sometimes, especially when generating feedback. I had to set longer timeouts and make sure the system told the user if it was thinking, plus put in good error handling. Keeping track of all the interview stages and making sure it flowed right (like if someone quit early or didn't answer) took a lot of testing and fixing. Getting the robot's gestures and speech to look natural together was tough at first; just making it do one then the other looked weird. I ended up using a special way to build the robot's output so it all happened smoothly, though it made the code a bit more complex. Also, relying on the AI to format its responses just right for the emotion tags was a bit risky, so I had to build in checks for that. And figuring out how long the robot should wait after the user paused before it spoke again needed a bit of trial and error to feel natural.

This project has some real-world uses, for sure. It can directly help people get ready for job interviews, which is a big deal with how competitive things are. Places like university career centres or hiring agencies could use it, or just anyone wanting to practice. Because you can set it up for different jobs and industries, it’s pretty flexible. What was learned about how much having a physical robot helps can also guide how new HR tech and training tools are made, especially for jobs where you need good people skills. It also shows a good way to link up smart AIs with social robots for more complicated, interactive training.

I learned a ton from doing this. I figured out how to actually build one of these Human-Robot Interaction things, connecting the FurHat robot to a smart AI like ChatGPT to make it feel like a real, changing conversation. The feedback from users really showed that having a physical robot makes a big difference to how real and engaging it feels, and even how well people spoke, compared to just a screen. I got a good handle on the nitty-gritty of making these systems work, things like dealing with API delays, writing good prompts for the AI, making the robot's speech and gestures sync up, and why setting up the turn-taking properly matters so much. It really hit home how tricky it is to design and test these interactive robot systems, especially when you’re using outside AI and trying to measure how people feel about it. One of the biggest things was seeing that even though the robot wasn't perfect, people still found it way more helpful to practice with than a virtual version. It showed that how you present an AI is a big deal, and a physical robot has some real advantages for making practice feel like the real thing. It also made it clear that while having a robot body is great, you’ve got to keep working on making its voice and movements super natural to get the full benefit.

Why did you choose this course?

I've always been pretty into tech and how things work, especially computers and software. I really like the problem-solving side of it, that feeling when you finally solve a problem or get a piece of code to do exactly what you want. Plus, Computer Science opens up a lot of doors career-wise, and I wanted a degree that would let me be involved in interesting projects and be part of how the future is shaping up.

What have you enjoyed most about your course so far?

What I've enjoyed most is probably getting to actually build things. It’s one thing to learn the theory, but it’s way more exciting to apply it in practical projects, whether that’s coding an application, working on a team to create something, or even just debugging a complex program until it finally clicks. Learning new programming languages and seeing how different technologies fit together has been really engaging too. There's always something new to figure out, which keeps it interesting.

What are the best things about studying your course?

The best thing is probably the power it gives you to create and innovate. You're learning the tools to bring ideas to life, which is incredibly rewarding. It’s a field that's constantly changing, so you're always learning, which is cool if you like to keep your brain busy. Plus, everything is so digital now, so what you learn feels really relevant to the world.

Computer Science gives you the building blocks to not just understand the digital world, but to actively shape it.

Adam Jones

BSc Computer Science

Why did you choose to study at Sheffield?

When I was looking at universities, Sheffield stood out for a few reasons. The School of Computer Science here has a really good reputation, and I liked the sound of the course structure, it seemed to cover a good range of topics I was interested in. Beyond the course itself, I'd heard great things about Sheffield as a student city. It felt like a place with a good vibe, friendly, and not overwhelmingly huge, but still with plenty going on.

What advice would you give to a younger student considering studying at Sheffield?

My main advice would be to make the most of everything once you're here. Explore the city properly. There are so many cool independent places and, of course, try to get out to the Peak District. Academically, really engage with your course, don't be afraid to ask lecturers or your course mates questions if you're stuck on something. And just try to find a good balance between your studies and enjoying student life.