The applications and potential of AI in science are enormous. AI not only helps to make existing knowledge more discoverable and more readily available. AI is also invaluable in the search for new materials, chemical compounds and biological agents, for example to use solar energy more efficiently or to develop drugs against emerging pathogens. In the process, scientists are faced with major challenges: They have almost infinite chemical and biological options. Such discoveries require end-to-end AI-powered automation – from experimental design to execution and analysis.

Today’s computational learning systems are not yet capable of realizing the full potential of AI-enabled materials, chemical, environmental, and life sciences. We need new AI methods that can both predict complex phenomena and provide insights into underlying processes. Such methods will be the basis for developing tailored systems capable of addressing major global health and environmental challenges. Research is being conducted at L3S on the necessary AI methods.  

5
5GAPS

5GAPS

5GAPS is a platform for real-time multi-attribute space capture and multi-dimensional object positioning aiming to digitally recreate Hanover in real-time
F
FoDaHemm

FoDaHemm

The FoDaHemm project aims to reduce barriers to data protection in research projects by examining the tension between the right to informational self-determination and freedom of research.
FID Pharmazie (SIS Pharmacy)

FID Pharmazie (SIS Pharmacy)

The SIS Pharmacy provides comprehensive access to subject-specific information resources and supports pharmaceutical scientists with customised services.
I
ixAutoML

ixAutoML

Making automatic machine learning systems more human-centered by enabling interactivity and explainability.
N
NFDI4 Data Science

NFDI4 Data Science

NFDI4 Data Science aims to establish a national research data infrastructure in Germany to support all stages of the interdisciplinary research data lifecycle for Data Science and Artificial Intelligence.
O
Online Optimierung

Online Optimierung

The goal of the project is to develop and investigate online convex optimization (OCO)-based control schemes for general cost functions and constraints without relying on restrictive assumptions.
P
PlanOS

PlanOS

The PlanOs project aims to develop large-area sensor networks in thin polymer films for strain and shape measurement with high efficiency, low cost and high resolution.
R
Robotics Lab

Robotics Lab

The L3S’s Robot Learning Lab works on foundational research to develop increasingly autonomous assistive systems.
Reflect AI

Reflect AI

The project aims to develop hybrid artificial intelligence models that use art-historical expertise and knowledge graphs to improve image similarity assessments in art history.
S
ScienceGRAPH

ScienceGRAPH

ScienceGRAPH aims to develop a novel principled model for representing scholarly communication in a knowledge-based way by using interlinked knowledge graphs.