Text
Smart Kitchens: How AI is Making Cooking Safer for the Elderly
2025-06-16
New AI research is making our homes safer
Researchers from Mälardalen University and the University of Greifswald have developed a cutting-edge system to enhance kitchen activity recognition, aimed at supporting elderly individuals and those with cognitive impairments. This innovative approach leverages advanced AI techniques to monitor and assist with daily kitchen tasks, ensuring safety and independence for vulnerable populations.
Understanding the Breakthrough:
As the global population ages, the demand for technologies that facilitate independent living, especially for those with cognitive impairments, is increasing. One of the critical areas where assistance is needed is in performing routine kitchen tasks, which can be complex and potentially dangerous for individuals with neurodegenerative disorders like Alzheimer's disease.
The system developed by the researchers can monitor kitchen activities in real-time, providing timely assistance and alerts to prevent accidents. This technology has the potential to revolutionize the field of assistive and remote care technologies, making daily living safer and more manageable for elderly individuals.
The team, including Samaneh Zolfaghari, Teodor Stoev, and Kristina Yordanova, has conducted a systematic evaluation and enhancement of the Rostock Kitchen Task Assessment (KTA) dataset. The researchers have refined the dataset, ensuring high-quality data for training AI models. They have also conducted extensive experiments to benchmark different activity recognition approaches, demonstrating superior results.
"Our goal is to provide reliable support for elderly individuals in their daily kitchen activities, ensuring they can live independently and safely," says Samaneh Zolfaghari.
How It Works:
The researchers have developed a prototype system that uses a combination of sensors and AI models to monitor and recognize kitchen activities. The system employs wearable sensors, embedded object sensors, and video data to gather information about the user's actions. This data is then processed using advanced machine learning techniques, including Hidden Markov Models (HMM) and Random Forest (RF), to accurately detect and classify kitchen activities.
One of the key innovations in this study is the combination of HMM and RF models, which has shown superior results in recognizing kitchen tasks. The system can identify common actions like pouring, stirring, and taking objects.
"By integrating multiple sensor technologies and advanced AI models, we can provide real-time assistance and monitoring, significantly improving the safety and independence of elderly individuals," notes Teodor Stoev.
Contributions to the UN Sustainable Development Goals (SDGs):
Goal 3: Good Health and Well-being
By improving the recognition of kitchen activities, this work supports safer, more independent living for elderly individuals and those with cognitive impairments—enhancing both well-being and quality of life.
Goal 9: Industry, Innovation and Infrastructure
The research drives the development of innovative, sensor-based technologies that enable sustainable, home-based care solutions and support independent living.
Bottom line: This research bridges the gap between technology and healthcare, providing practical solutions for everyday challenges faced by the elderly.