Select Page

PlantGrid (2020-2023)

funded by Federal Office of Food and Agriculture

We are conducting research in several funded projects. You can find more information about the project PlantGrid below. If you are interested in more details and discussions about our projects, do not hesitate to contact us.
For more information please contact Dominik Grimm.

Project Description

The demand for horticultural products is depending on various external factors making it hard to predict sales. In practical operation, this may lead to out-of-stock situations possibly causing missed sales as well as excess-stock cases, which can result in the destruction of products. Besides financial loss, the latter is also generating environmental damage due to wasted resources during production and transport. These matters encourage conducting research in this area. Hence, the research project Digital management support systems for small and medium-sized enterprises in value chains of ornamental plants, perennials and cut flowers (acronym: PlantGrid) was initiated. The goal of this project is to support small and medium-sized horticultural companies with a digital management system. This system is supposed to improve planning and disposition of goods, e.g. in order to reduce out-of-stock and excess-stock situations. Moreover, we will take businesses along the whole horticultural value chain starting with producers leading to wholesalers and ending with retailers into account.

Besides several horticultural companies and the digital agency snoopmedia, PlantGrid is done in cooperation with the professorships Marketing and Management of Biogenic Resources and Retail Management of the University of Applied Sciences Weihenstephan-Triesdorf as well as the professorships Logistics Management and Horticultural Economics of the University of Applied Sciences Geisenheim.

One essential part of this project are Multivariate Time Series Predictions using historical observations and external factors. This part is the main focus of the bioinformatics lab within the project. At the beginning, we will examine internal and external data sources potentially relevant for predictions in horticulture. First, we will explore the retail sector, and evaluate models using preliminary datasets of a selection of companies as well as external data sources. Initially, we will apply methods like Exponential Smoothing and ARIMA or its extensions, which serve as baselines and comparison partners. Afterwards, modern ML-based methods will be evaluated, especially in a multivariate context. Finally, we will examine the combination of different predictors using simple and performance-based techniques. The steps described above should lead to a robust and precise prediction model for the horticultural retail sector. Next, we will enhance the results with additional information along the value chain. Furthermore, handling data heterogeneity and insufficiency are interesting research directions for the proposed project and therefore a central point for the bioinformatics lab.

You can find more information in German here.

Involved people in our team:




ForeTiS: A comprehensive time series forecasting framework in Python.
J Eiglsperger*, F Haselbeck*, DG Grimm (* equal contribution)
Machine Learning with Applications, Vol.12, 2023
( [Code] [Documentation]

Dynamically Self-Adjusting Gaussian Processes for Data Stream Modelling.
JD Hüwel*, F Haselbeck*, DG Grimm, C Beecks (* equal contribution)
KI 2022: Advances in Artificial Intelligence, 2022
( [Code]

EVARS-GPR: EVent-triggered Augmented Refitting of Gaussian Process Regression for Seasonal Data
F Haselbeck, DG Grimm
KI 2021: Advances in Artificial Intelligence, 2021
( [Code]

Machine Learning Outperforms Classical Forecasting on Horticultural Sales Predictions
F Haselbeck, J Killinger, K Menrad, T Hannus, DG Grimm
Machine Learning with Applications, Vol.7, 2022
( [Code]



ForeTiS: A Forecasting Time Series Framework

Time series forecasting is a research area with applications in various domains, nevertheless without yielding a predominant method so far. We present ForeTiS, a comprehensive and open source Python framework that allows rigorous training, comparison, and analysis of state-of-the-art time series forecasting approaches. Our framework includes fully automated yet configurable data preprocessing and feature engineering. In addition, we use advanced Bayesian optimization for automatic hyperparameter search. ForeTiS is easy to use, even for non-programmers, requiring only a single line of code to apply state-of-the-art time series forecasting. Various prediction models, ranging from classical forecasting approaches to machine learning techniques and deep learning architectures, are already integrated. More importantly, as a key benefit for researchers aiming to develop new forecasting models, ForeTiS is designed to allow for rapid integration and fair benchmarking in a reliable framework. Thus, we provide a powerful framework for both end users and forecasting experts.

ForeTiS is available at We provide a setup using Docker, as well as a Python package at

Extensive online documentation with hands-on tutorials and videos can be found at


More information can also be found in the following publication. Please cite our publication when using ForeTiS.

ForeTiS: A comprehensive time series forecasting framework in Python.
J Eiglsperger*, F Haselbeck*, DG Grimm (* equal contribution)
Machine Learning with Applications, Vol.12, 2023



Talk on EVARS-GPR at 44th German AI Conference



Professorship Bioinformatics

Petersgasse 18
94315 Straubing


Prof. Dr. Dominik Grimm

Phone: +49 (0) 9421 187-230


Team Assistants

Anna Fischer (Maternity Leave)

Jasmin Schneider

Phone: +49 (0) 9421 187-201