Research

In an important study for understanding how memories are made, EPFL scientists show that the flexibility of chromatin – packaged DNA inside the cell – plays a crucial role in “deciding” which neurons are involved in forming a specific memory.

To transition away from fossil fuels and towards renewables – which are intermittent by nature – we’ll need to rework our entire system of power storage, transmission and distribution. Yet experts haven’t yet found the right energy mix or power storage system, or how to balance supply and demand effectively.

Under the leadership of Empa scientist Mashael Yazdanie, an international research team is investigating ways to better plan for climate-resilient energy systems in the Global South. Focusing on the case study of Accra, the capital of Ghana, the multidisciplinary team expanded conventional energy system modeling approaches by incorporating a range of socio-techno-economic challenges, climate change impacts, and resilience metrics into their models. Their approaches are applicable worldwide to support widespread sustainable and resilient energy system transitions.

Data collected by satellites, drones, radars and microscopes provide a goldmine of information to better understand our environment. And when these data are coupled with artificial intelligence (AI), they can unlock the secrets of phenomena taking place at all levels. 

A new study by the Swiss Academy of Engineering Sciences (SATW) shows how artificial intelligence (AI) can be used to reduce the impact of climate change on Switzerland. The experts involved have high hopes for satellite-based earth observations and AI models. According to Empa researcher Gerrit Kuhlmann, one of the co-authors of the study, these could quickly identify the sources of greenhouse gas emissions.

Addressing problems with diagnosing and treating breast cancer, scientists at EPFL have developed EMBER, a tool that integrates breast cancer transcriptomic data from multiple databases. EMBER can improve precision oncology by accurately predicting molecular subtypes and therapy responses.

Researchers at Eawag and the Swiss Data Science Center have trained AI algorithms with a comprehensive ecotoxicological dataset. Now their machine learning models can predict how toxic chemicals are to fish.

Developed at EPFL, Heriot-Watt University, and ETH Zurich, PrISMa is a new platform that uses advanced simulations and machine learning to streamline carbon capture technologies, by taking into account the perspectives of diverse stakeholders early in the research process.

EPFL was selected by the EUROfusion consortium to develop an advanced visualization system for a preliminary process in nuclear fusion. The system is designed to turn reams of simulation and testing data into real-time graphics worthy of today’s advanced video games.

Researchers from ETH Zurich and Empa have developed a hydrogel implant that can help prevent endometriosis, a condition that affects a great many women. This innovation also acts as a contraceptive.

EPFL researchers discovered that deepwater renewal in Lake Geneva in wintertime is not only due to vertical mixing. Instead, strong currents coming from the lake’s Petit Lac basin and nearshore zones of the Grand Lac play a vital role.

When the Earth’s ice masses melt, the way the planet rotates also changes. Researchers at ETH Zurich have now been able to show how climate change is altering the Earth’s axis of rotation and the length of the day. The speed of rotation, which was hitherto mainly influenced by the moon, will now also depend much more on the climate.

A newly developed airshield supports track and field athletes such as Mujinga Kambundji with overspeed training.

Biodiversity is dwindling rapidly, while measures to protect it are lagging behind the targets. Only one thing can help: utilising scarce resources as efficiently as possible where they are most effective. The “Lanat-3” research project lays the foundation for this with the help of the latest data and AI-supported models.

Researchers at EPFL have made a breakthrough in understanding how neural network-based generative models perform against traditional data sampling techniques in complex systems, unveiling both challenges and opportunities for AI's future in data generation.