Ericsson has introduced Cognitive Labs, a virtual research initiative dedicated to advancing artificial intelligence (AI) technologies such as Graph Neural Networks (GNNs), Active Learning, and Large-Scale Language Models (LLMs). The program operates with teams in Madrid, Málaga, and Cairo. Focus areas include enhancing mobile communications and exploring emerging technology applications in sectors like healthcare. The press release notes that the lab will collaborate with the open-source community and develop more effective AI projects for its customers and academic partners.
[Ericsson]
- GAI Lab (Geometric Artificial Intelligence Lab)
Concentrates on Geometric AI, including geometric learning, graph generation, and temporal GNNs. This research holds promise for fields such as drug discovery. - MLR Lab (Machine Learning and Reasoning Lab)
Aims to optimize model training and reinforcement learning, reducing both time and energy consumption. This lab also explores digital twins—virtual simulations of real-world systems—to improve efficiency across various industries. - FAI Lab (Fundamental Artificial Intelligence Lab)
Focuses on foundational AI models, particularly LLMs, to automate processes and enhance telecommunications. By refining language models, the lab seeks to bolster communication networks.
The lab’s mission is to “create world-class research that benefits the academic community and enhances our products to serve our customers better,” according to Jean-Christophe Laneri, Head of Cognitive Network Solutions at Ericsson.
In Spain, Ericsson’s R&D activities in Cognitive Network Solutions (CNS) are based in Madrid and Málaga, making the region a focal point for Cognitive Labs. Recently, Ericsson Spain and Universidad Pontificia Comillas (ICAI) signed a research alliance to collaborate on open-source projects and co-author publications. This pact unifies Ericsson’s data scientists and ICAI’s engineering teams.
