Can AI revolutionize elderly care? Professor Marco Spruit, an expert in translational data science, is developing Well-being AI—a smart pet robot that provides companionship while monitoring health through language markers. This technology could ease pressure on healthcare services by detecting early signs of decline and alerting caregivers.

Beyond professional care, these intelligent robots could also support informal caregivers. Imagine a situation where one partner begins to suffer from dementia—this technology would allow the other partner to remotely monitor their well-being, offering peace of mind and timely intervention. This way AI can be the key to a more sustainable and compassionate future for elderly care?  

How can nature-inspired evolution help us create the perfect optical lens? Assistant Professor Anna Kononova explores this question in her research. High-tech industries, like computer chip manufacturers, require ultra-precise lenses to keep up with the demand for increasingly complex designs.

Anna uses evolutionary computing, a technique that mimics natural selection. Just as nature evolves the fittest species, computers generate and test many lens designs, keeping the best and refining them over generations. This approach helps create cutting-edge technology with incredible precision. 

Ever wondered how algorithms make sense of complex problems? In this episode, we dive deep into the art of designing algorithms with associate professor Mitra Baratchi. Exploring how computers process data, find connections, and deliver insights.

While algorithms can solve many challenges, there are limits to their accuracy. How can researchers identify these gaps and guide users toward the best solutions. How much should we rely on algorithms, and where does human expertise come in? 

How and why do humans form relationships with AI? This question lies at the core of Assistant Professor Peter van der Putten’s research. While AI sparks fear—raising concerns about its impact on our place in the universe—it also fascinates us, offering glimpses into an unknown future.

But can we truly bond with AI? It’s one thing to feel attached to a humanoid robot or a robotic pet, but what about a simple wooden cube with a moving face? In his study Common Locus, Peter explored just that. The results might surprise you—and reveal more about human nature than you’d expect.

If you want to read more about the research project of Joost Mollen, Peter van der Putten, and Kate Darling: Bonding with a Couchsurfing Robot: The Impact of Common Locus on Human-Robot Bonding In-the-wild. ACM Transactions on Human-Robot Interaction. 12, 1, Article 8, March 2023.

More information about Peter can be found here: https://liacs.leidenuniv.nl/~puttenpwhvander/ 

In this episode, LIACS Assistant Professor Hao Wang shares insights into multi-criteria optimization and the concept of the Pareto optimum. Discover how these powerful techniques, often used in decision-making and problem-solving, can help you plan the perfect trip.

Whether you’re balancing cost, travel time, and comfort or choosing between destinations, learn how mathematical models can guide you toward the best possible choices. Tune in for a new episode that bridges research and real-life experiences.

The hay fever season lasts longer than it used to because of climate change. LIACS Assistant Professor Lu Cao explains how machine learning and AI can help to easier count pollen in the air. This improves the accuracy and efficiency of pollen monitoring, a critical tool for understanding seasonal allergies, air quality, and climate impact.

Discover how AI-driven technologies can improve the accuracy and efficiency of pollen monitoring, providing valuable insights for municipal governments. With better data, policymakers can adjust their strategies on public green spaces, balancing ecological benefits with the needs of residents affected by allergies. 

Can robots feel joy or guilt? Can they have emotions at all? Today, we dive into the fascinating world of "affective" computing with Associate Professor Joost Broekens, a leading expert in the field. But what exactly is affective computing?

We also tackle the ethical challenges of AI-emotion interaction, exploring topics like the European AI Act, which bans emotion recognition in education and the workplace. Are these new technologies transforming artificial intelligence into emotional intelligence?

Professor Vedran Dunjko leads the Applied Quantum Algorithms group at Leiden University. His research focuses on quantum computing, which uses quantum effects to greatly improve computation. This could lead to much more powerful computers. Developing a working quantum computer is an ongoing effort, bringing both exciting possibilities and potential security risks because of its superior computing abilities.

Much of what a quantum computer can achieve remains a mystery, even to experts like Dunjko. Essentially, the quantum computer is a solution in search of a problem. Nonetheless, Dunjko is confident in its potential applications, particularly in the field of quantum physics.  

Assistant Professor Hazel Doughty specializes in computer vision, aiming to enable computers to automatically understand visual content. Her research focuses on video understanding, where thousands of videos with text descriptions are fed to the computer. Her project, "From What to How," seeks to advance computers from identifying actions to understanding how they are performed, though this is currently limited by insufficient data.

Ultimately, her goal is for computers to independently analyze videos and comprehend methods used in various tasks. This would allow for the creation of assistive videos for activities like for instance proper CPR techniques.

Assistant Professor Jan van Rijn is at the forefront of research into the trustworthiness of AI systems. His work focuses on the field of machine learning, with a particular emphasis on exploring and enhancing the robustness of AI systems. Robustness is a critical component in building trustworthy machines, such as self-driving cars, where reliable performance is paramount.

Jan and his team are dedicated to developing techniques that make verification models for testing robustness more efficient. However, significant challenges remain. One of the primary obstacles is the sheer size of current AI systems, like ChatGPT, which are too large to be efficiently verified using existing methods.