A Finnish Scientist: “Discovering How Humans and Robots Can Safely Work Together Might Spark a New Industrial Revolution”

The intersections of cognitive robotics, neuroscience, and artificial intelligence (AI) have long captured humanity’s imagination, offering glimpses of an exciting and transformative future. Dr Roel Pieters, Associate Professor at Tampere University (TAU) in Finland, has been working in the Cognitive Robotics research group for eight years. He is also contributing to the AI Competence Centre project SustAInLivWork in Lithuania. Together with his colleagues, Dr Pieters focuses on advancing human-robot interaction and envisions a new industrial breakthrough driven by robots that can safely collaborate with humans – robots capable of understanding commands through simple gestures or spoken words.

Dr Roel Pieters
Dr Roel Pieters

How and why was the Cognitive Robotics group established at your university, and what research is it focused on?

I joined Tampere University (TAU) in 2017 as an associate professor and started the Cognitive Robotics research group. Currently, the group has around ten people from different backgrounds and in different positions, from student assistants to doctoral students and postdoctoral researchers. We participate in several European and national research projects related to robotics and AI applications in industries such as manufacturing and inspection and maintenance. Naturally, we also design and deliver educational programmes (both bachelor’s and master’s levels) focused on robotics and supervise student thesis projects.

One of our main research goals is enabling robots to safely and efficiently work alongside humans – a development that could truly revolutionise industries. Currently, most robots, especially in industrial manufacturing, have little to no interaction with humans. Factories strictly separate robots and humans using safety fences, and robots are programmed to stop moving when a human gets too close.

Our group initially focused on programming and ensuring safety for such robots in industrial environments. Over time, we integrated advanced AI technologies, such as perception tools and large language models (LLMs), to expand the capabilities and usability of robots in these settings.

Put simply, we are searching for ways for humans and robots to work together safely, with robots acting as assistants that take on physically demanding or dangerous tasks. Communication is a key aspect of this collaboration – humans must be able to easily and accurately convey what the robot needs to do and when.

What are the main challenges in teaching robots to better understand humans?

Programming robots and ensuring reliable communication remain significant challenges. While AI helps with perception and decision-making, it is still limited in many ways. This means humans must often undergo extensive training to use these systems effectively, and safety remains the top priority.

That said, we’ve made considerable progress through various projects. For example, we’ve demonstrated how humans can communicate with robots using different modalities – speech, gestures – and how AI tools can enhance a robot’s situational awareness when collaborating with humans.

Dr Roel Pieters at SustAinLivWork annual meeting

You’ve been working with the Cognitive Robotics group for eight years now. What keeps you passionate about this project?

Two things. First, it’s incredibly inspiring to work with students – their skills and motivation never cease to amaze me!

Second, participating in large-scale European projects like OpenDR has been very rewarding. In OpenDR, we developed open deep-learning tools for robotics. Such collaborations between researchers from different countries enable us to create innovations with broad applications.

Overall, I believe universities and researchers play a vital role in advancing robotics and AI. Initiatives like the AI Competence Centre SustAInLivWork in Lithuania, provide a platform where experts from diverse fields can pool their knowledge and research. These efforts not only foster innovation but also bridge the gap between academia and industry, enabling businesses to adapt and monetise research outcomes.

Tampere Robolab
Tampere Robolab

Which area of cognitive robotics excites you the most and seems especially promising?

At the moment, I am particularly fascinated by visual-language-action models. These AI models combine visual information with text or speech to generate robot commands without requiring manual programming.

In other words, such technologies would allow robots to perform tasks by simply interpreting verbal commands from humans. You say what you want, and the robot executes it.

Cognitive robotics also benefits from advances in related fields, such as artificial neural networks. These networks allow robots to learn from data, recognise patterns, and make decisions, significantly expanding their capabilities and adaptability. I believe we’re still only scratching the surface of what AI can achieve, especially in human-robot interaction.

Cognitive robotics often captures public attention due to its ties with AI, neuroscience, and the future of human-robot collaboration. What do you see as its potential in everyday life? How could it transform industries such as healthcare, education, and logistics?

Cognitive robotics is already making a significant impact in various industries. In healthcare, robots assist in surgeries and patient care, improving precision and efficiency. In education, they serve as teaching aids, making learning more interactive and engaging. In logistics, robots optimise warehouse operations and delivery processes, boosting productivity and reducing costs.

So, while we don’t yet see robots directly in our everyday lives, their applications in these industries bring tangible benefits to both businesses and society.

The concept of human-robot interaction has long fascinated writers and filmmakers, leading to many myths and misconceptions. What are some of the most common misunderstandings about robots today?

One common misconception is overestimating what robots can do. While robots are becoming more advanced, they still have many limitations compared to humans. Certain tasks are physically impossible for robots, or they are too complex for humans to programme.

Another frequent misunderstanding is assigning personality traits to robots, which can lead to unrealistic expectations about their capabilities and interactions.

Finally, you work with many young, curious people from around the world. What advice would you give to aspiring researchers interested in cognitive robotics and AI?

Get started and be hands-on! Most importantly, seize any opportunity to work with real robots. For instance, at Tampere Robolab, we provide students with access to our robots. Alternatively, use simulation packages to familiarise yourself with robotics. At our university, we use ROS (Robot Operating System), which offers all the tools needed to become an expert in this field. Additionally, you can expand robot functionalities by integrating other AI tools.

The project is co-funded under the European Union's Horizon Europe programme under Grant Agreement No. 101059903 and under the European Union Funds’ Investments 2021-2027 (project No. 10-042-P-0001).

© 2025 SustAInLivWork.eu. All Rights Reserved.