Logo des Repositoriums

Auflistung nach Autor:in "Guldner, Achim"

1 - 10 von 10
  • Konferenzbeitrag
    An Artificial Intelligence of Things based Method for Early Detection of Bark Beetle Infested Trees
    (EnviroInfo 2022, 2022) Knebel, Peter; Appold, Christian; Guldner, Achim; Horbach, Marius; Juncker, Yasmin; Müller, Simon; Matheis, Alfons
    Bark beetles, like the European Spruce Bark Beetle (Ips typographus), are inherent partsof a forest ecosystem. However, with favorable conditions, they can multiply quickly and infest vastamounts of trees and cause their extinction. Therefore, it is important for forest officials and rangers ofe. g. a national park, to monitor the population of the beetles and the infested trees. There are severalways to approach this, but they are often costly and time-consuming. Therefore, we design and test abark beetle early warning system with AI-based data analysis: Audio data, data on pheromones andinformation for a drought stress assessment of the affected trees are to be collected and used as a basisfor the analysis. The aim is to devise a micro-controller-based sensor system that detects the infestationof a tree as early as possible and warns the forest officials, e. g. via a message on their cell phone.
  • Konferenzbeitrag
    Detecting Consumer Devices by Applying Pattern Recognition to Smart Meter Signals
    (Proceedings of the 27th Conference on Environmental Informatics - Informatics for Environmental Protection, Sustainable Development and Risk Management, 2013) Guldner, Achim; Arns, Sebastian; Schunk, Tobias; Gollmer, Klaus-Uwe; Michels, Rainer; Naumann, Stefan
    Future energy supply requires an intelligent load management for efficient distribution of the available energy, at national level, as well as on a regional scale. For this purpose, one necessary prerequisite is the immediate detection of the currently connected appliances (loads), for example white or brown goods. If the devices that are currently active at the time of a data point are known, it is possible to level the load curve by means of selectively connecting and disconnecting appliances, which results in an optimized usage of the available energy. To realize the measurement of the energy consumption, we devised a low-investment system for centralized data acquisition and recorded and digitized characteristic load profiles. Afterwards, the application of different pattern matching algorithms allowed for recognizing and assigning individual loads from the measured sum signal. In the course of laboratory experiments, we could identify individual appliances and their combinations with this system.
  • Konferenzbeitrag
    Development of a Real-time Smart Meter for Non-Intrusive Load Monitoring and Appliance Disaggregation
    (EnviroInfo & ICT4S, Adjunct Proceedings, 2015) Jonetzko, Roman; Detzler, Matthias; Gollmer, Klaus-Uwe; Guldner, Achim; Huber, Marcel; Michels, Rainer; Naumann, Stefan; Ney, Martin
    Current studies about electrical energy efficiency potentials bring out that to make the most part of the saving potentials achievable, a feedback about the instantaneous consumed electrical energy is necessary. More detailed, an allocation of the electrical load to the particular device can enable a greater level of sensitivity in energy consumption of electrical appliances. Therefore, we pursued the aim of developing a low cost smart meter hardware, which fits the requirements for the detection of several devices and states via disaggregation algorithms. This is implemented by applying Fast Fourier Transformation to the measured data and sending the Fourier coefficients to the appliance disaggregation modul. In this paper, we describe the developed hardware in detail and show a visualization approach of the disaggregation results for providing the user with detailed information about device states.
  • Textdokument
    Exploration and systematic assessment of the resource efficiency of Machine Learning
    (INFORMATIK 2021, 2021) Guldner, Achim; Kreten, Sandro; Naumann, Stefan
    Estimations of today’s energy consumption of information and communication technologies (ICT) range from 2 to 9 % of the total produced energy and forecasts for the year 2030 predict an increase up to 21 %. Even though these numbers are controversial, it cannot be denied that the consumption growth of large impact factors, like data centers, networks, consumer devices, and the production of ICT needs to be reduced. In addition to Green IT, which is primarily focused on hardware, software is increasingly seen as an energy consumer with considerable savings potential. In this paper, we take a look at software for artificial intelligence (AI) and especially machine learning (ML). We describe a method for in-depth measurement and analyses of the energy consumption and hardware usage of ML algorithms and a series of experiments where we use the method on convolutional neural networks (CNN). We also compare existing estimation methods with our own. As outlook, we propose a holistic approach along the AI life cycle and additional experiments and assessments that could show potential efficiency improvements and consumption savings in AI.
  • Konferenzbeitrag
    Greening software with continuous energy efficiency measurement
    (INFORMATIK 2013 – Informatik angepasst an Mensch, Organisation und Umwelt, 2013) Drangmeister, Jakob; Kern, Eva; Dick, Markus; Naumann, Stefan; Sparmann, Gisela; Guldner, Achim
    The energy consumption of information and communication technology (ICT) is still increasing. Several solutions regarding the hardware side of Green IT exist, until now the software contribution to Green IT is not considered sufficiently apart from scientific research. In our paper, we discuss a new method of improving the energy efficiency of software during its development, by putting energy efficiency measurements into practice. Therefore, we measure and rate energy consumption and efficiency during the software development process, based upon software testing and Continuous Integration (CI).
  • Konferenzbeitrag
    Integrating Aspects of Carbon Footprints and Continuous Energy Efficiency Measurements into Green and Sustainable Software Engineering
    (Proceedings of the 27th Conference on Environmental Informatics - Informatics for Environmental Protection, Sustainable Development and Risk Management, 2013) Kern, Eva; Dick, Markus; Drangmeister, Jakob; Hiller, Tim; Naumann, Stefan; Guldner, Achim
    The energy consumption of information and communication technology (ICT) is still increasing. Even though, up to now, several solutions regarding the hardware side of Green IT exist, the software contribution to Green IT is not well investigated. In our paper, we discuss how to integrate some aspects of carbon footprint calculation into software development processes and we show how ongoing energy efficiency measurements can be established as an integral part of a software development project.
  • Konferenzbeitrag
    Potentials of Green Coding - Findings and Recommendations for Industry, Education and Science
    (INFORMATIK 2023 - Designing Futures: Zukünfte gestalten, 2023) Junger, Dennis; Westing, Max; Freitag, Christopher; Guldner, Achim; Mittelbach, Konstantin; Weber, Sebastian; Naumann, Stefan; Wohlgemuth, Volker
    Progressing digitalization and increasing demand and use of software cause rises in energy- and resource consumption from information and communication technologies (ICT). This raises the issue of sustainability in ICT, which increasingly includes the sustainability of the software products themselves and the art of creating sustainable software. To this end, we conducted an analysis to gather and present existing literature on three research questions relating to the production of ecologically sustainable software (’Green Coding’) and to provide orientation for stakeholders approaching the subject. We compile the approaches to Green Coding and Green Software Engineering (GSE) that have been published since 2010. Furthermore, we considered ways to integrate the findings into existing industrial processes and higher education curricula to influence future development in an environmentally friendly way.
  • Konferenzbeitrag
    Sustainability in Artificial Intelligence - Towards a Green AI Reference Model
    (INFORMATIK 2023 - Designing Futures: Zukünfte gestalten, 2023) Weber, Sebastian; Guldner, Achim; Begic Fazlic, Lejla; Dartmann, Guido; Naumann, Stefan
    The interest in Green Artificial Intelligence (AI) is growing as AI research is increasingly focusing on and taking into account environmental sustainability. This paper aims to clarify and emphasize the distinction between terms like sustainable AI, Green AI, Green by AI, and Green in AI, highlighting their importance in the context of environmentally responsible AI practices. We find that existing Green Software reference models are insufficient for meeting the unique requirements of Green AI. Thus, we argue that a tailored Green AI reference model is needed to guide and promote environmentally responsible practices in the field of AI, addressing the special considerations associated with Green AI.
  • Konferenzbeitrag
    Towards Sustainable Machine Learning: Analyzing Energy-Efficient Algorithmic Strategies for Environmental Sensor Data
    (INFORMATIK 2024, 2024) Cetkin, Berkay; Begic Fazlic, Lejla; Guldner, Achim; Naumann, Stefan; Dartmann, Guido
    This study evaluates the energy efficiency of machine learning (ML) classification models across 49 test setups, each representing different conditions derived from a set of scenarios. Utilizing internet of things (IoT) technology with an ESP8266 microcontroller, we collected and analyzed environmental data including temperature, humidity, and CO2 levels from a simulated room environment. We measured energy consumption for data preprocessing, model training, and testing, alongside energy efficiency metrics that consider output, processing time, and F1 score. The study also performed correlation analyses to explore the relationship between energy consumption and performance metrics. Furthermore, it assessed the trade-offs between accuracy and energy efficiency by comparing an ensemble model to its constituent algorithms. The measurements, conducted according to the Green Software Measurement Model (GSMM), provide essential insights into selecting energy-efficient algorithms for a broad spectrum of IoT applications.
  • Konferenzbeitrag
    Was weiß ChatGPT über Nachhaltige Software-Entwicklung und Green Coding? Erste Tests und Bewertungen
    (INFORMATIK 2023 - Designing Futures: Zukünfte gestalten, 2023) Naumann, Stefan; Guldner, Achim; Weber, Sebastian; Westing, Max
    In den letzten Jahren sind Large Language Models (LLM) wie GPT, BERT, PaLM, BLOOM oder LLamA durch die Verbesserung der Generierung natürlich wirkender Texte sowie neuer Fähigkeiten (Generierung von Quellcode etc.) stark ins öffentliche Interesse gerückt. Durch den Black Box-Charakter der Modelle und der Unklarheit der Qualität der zugrundeliegenden Trainingsdaten ist die Korrektheit der LLM-generierten Texte jedoch, insbesondere im wissenschaftlichen Umfeld, unklar und bedarf der Bewertung durch Expert:innen. In diesem Beitrag gehen wir daher auf die Frage ein, inwieweit LLM und insbesondere ChatBots wie ChatGPT Themen wie Green Coding und Nachhaltige Software-Entwicklung unterstützen können, und wie zuverlässig die Antworten am Beispiel von ChatGPT (GPT-3.5) sind. Wir stellen fest, dass die Inhalte für einen Überblick über die Themen sowie einen Einstieg in das Thema durchaus nutzbar sind, jedoch insbesondere im Hinblick auf weiterführende Quellen Schwachstellen aufweisen.