Advanced AI solutions
for Space Research and Technology
Bias Variance Labs, d.o.o. (BVLabs) is a research-led company, established by experts with long-standing experience in designing, developing and applying cutting-edge AI solutions for modeling and analysis of complex systems in space research and technology. The core focus of BVLabs is to facilitate the data-to-discovery & decision process by developing state-of-the-art explainable approaches that leverage latest developments in Artificial Intelligence and Data Science.
BVLabs provides novel solutions that address all aspects of the data lifecycle: Data Storage and Stewardship, Data and Knowledge Representation, Machine Learning, Data Analysis and Visualization. The team has complementary skills, addressing different areas such as: knowledge representation, design and application of efficient and interpretable learning approaches as well as modeling and analyzing complex and dynamic data. This blend of the expertise provides a unique and holistic view on the data-to-discovery and decision process leading to successful development and deployment of applications in the space research and technology sectors.
Projects
AI Toolbox for Earth Observation
AiTLAS
Contractor: ESA/ESRIN
AiTLAS focuses on designing and developing a platform that implements state-of-the-art AI methods for analysis of satellite images. The utility of AiTLAS will be demonstrated in three use-cases: (1) Benchmark repository of datasets, (2) Maya archaeological sites location, and (3) crop type prediction for Slovenia, Netherlands and Denmark.
Machine Learning for Spacecraft Operations
GalaxAI
Contractor: ESA/ESOC
GalaxAI focuses on designing and developing software prototypes that implement state-of-the-art machine learning methods and follow the best machine learning practices in order to deliver prototypes for Mars Express (MEX) Thermal Power Consumption Forecast and Integral Radiation Belts Crossing Prediction.
Value estimation of used products
Valeri
Contractor: Eligma d.o.o
We designed and developed machine learning methods that accurately predict the price of used products. These methods are part of the Eligma platform.
Contact
Address
Bias Variance Labs, d.o.o.
Koprska ulica 72,
1000 Ljubljana,
Slovenia