Henry Ekwaro-Osire

Mr. Henry Ekwaro-Osire

ICT Applications for Production, BIBA - Bremer Institut für Produktion und Logistik GmbH, Bremen, Germany
orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
7 January 2025 in Industrial Artificial Intelligence
Verified
Journal Article
Open access
PDF file
Henry Ekwaro-OsireHenry Ekwaro-Osire
Henry Ekwaro-Osire
SPSai Lalitha Ponugupati
Sai Lalitha Ponugupati
ANAbdullah Al Noman
Abdullah Al Noman
DBDennis Bode
Dennis Bode
KTKlaus-Dieter Thoben
Klaus-Dieter Thoben
Over the past two decades, machine learning (ML) has transformed manufacturing, particularly in optimizing production and quality control. A significant challenge in ML applications is obtaining sufficient training data, which data augmentation aims to address. While widely applied to image, text, and sound data, data augmentation for numerical data in manufacturing has seen limited investigation. This paper empirically compares three data augmentation techniques—generative adversarial networks, variational auto-encoders mixed with long-short-term memory, and warping—on four manufacturing datasets. It also provides a literature review, highlighting that generative models are the most common technique for numerical manufacturing data. Preliminary findings suggest that generative adversarial networks are effective for non-time-series numerical data, especially with datasets featuring many correlated model features, multiple machines, and sufficient instances and labels. This research enhances the understanding of data augmentation in manufacturing ML applications, emphasizing the need for tailored strategies.

24 November 2022 in Sustainability
Verified
Journal Article
Open access
PDF file
Henry Ekwaro-OsireHenry Ekwaro-Osire
Henry Ekwaro-Osire
DBDennis Bode
Dennis Bode
KTKlaus-Dieter Thoben
Klaus-Dieter Thoben
JOJan-Hendrik Ohlendorf
Jan-Hendrik Ohlendorf
Machine learning (ML) can be a valuable tool for discovering opportunities to save energy and resources in manufacturing systems. However, the hype around ML in the context of Industry 4.0 in the past few years has led to blind usage of the approach, occasionally resulting in usage when another analysis approach would be better suited. The research presented here uses a novel matrix approach to address this lack of differentiation of when to best use ML for improving energy and resource efficiency in manufacturing, by systematically identifying situations in which ML is well suited. Seventeen generic levers for improving manufacturing energy and resource efficiency are defined. Next, a generic list of six manufacturing data scenarios for when ML is a good method of choice for analysis is created. This results in a comprehensive matrix in which each lever is evaluated along each ML scenario and given a point, providing a quantitative ML suitability score for each lever. The evaluation is conducted by drawing on past studies demonstrating whether ML is appropriate. Specifically, operation parameter and input material optimization, as well as intelligent maintenance, are the levers that score the highest and are thus identified to be most suitable for machine learning. The majority of the remaining levers is deemed to have low suitability for machine learning. This simple yet informative matrix can be used as a guideline in data-driven manufacturing energy and resource efficiency projects to provide an appraisal on the applicability of ML as the initial analysis tool of choice.

20 November 2024 in Journal of Intelligent Manufacturing
Verified
Journal Article
Open access
PDF file
Henry Ekwaro-OsireHenry Ekwaro-Osire
Henry Ekwaro-Osire
DBDennis Bode
Dennis Bode
JOJan-Hendrik Ohlendorf
Jan-Hendrik Ohlendorf
KTKlaus-Dieter Thoben
Klaus-Dieter Thoben
This paper investigates the appropriateness of machine learning (ML) and other techniques for modeling manufacturing processes energy consumption by developing a comparison methodology. Three research questions are posed: Firstly, how do prediction errors compare using different techniques with varying complexity and ML use? Secondly, how does performance vary with different amounts of data? Thirdly, how do different techniques compare in terms of required expertise, effort to build and interpretability of results? To answer these questions, the authors develop a structured approach, which is also envisioned to be useable by practicing engineers and manufacturers. Four modeling categories are defined, ranging from simple non-ML methods, such as linear regression, to complex ML methods, such as deep neural networks. The approach is evaluated using data from a compound feed manufacturing process. The results confirm the notion that non-ML models are better suited to understand and model manufacturing processes when few parameters are present, due to their high interpretability, while ML models are recommended for analyzing processes with many potentially relevant and interrelated parameters. Interestingly the approach finds that the complex ML category model does not outperform the simple ML category model in terms of prediction accuracy, and only has the drawback of requiring more expertise to build and having lower interpretability. The study concludes that the decision to use complex ML for modeling manufacturing process energy consumption should be critically questioned and that a simpler approach may be better suited, suggesting that the developed methodology would be of value to practicing engineers.

31 August 2021
Verified
Book Chapter
Closed access
PDF file
Henry Ekwaro-OsireHenry Ekwaro-Osire
Henry Ekwaro-Osire
SWStefan Wiesner
Stefan Wiesner
KTKlaus-Dieter Thoben
Klaus-Dieter Thoben

26 April 2023
Verified
Book Chapter
Open access
PDF file
MKMichel Kaufmann
Michel Kaufmann
Henry Ekwaro-OsireHenry Ekwaro-Osire
Henry Ekwaro-Osire
KTKlaus-Dieter Thoben
Klaus-Dieter Thoben
Various sustainability reporting tools (SRTs) already exist, such as ISO 14001, Carbon Disclosure Project (CDP), and Greenhouse Gas Protocol (GGP). However, these tools focus on calculating total environmental impact for reporting to external parties, rather than providing guidance to a company on obtaining measures needed to identify improvement potentials within its production. This is due to low granularity of data measurements dictated by the SRTs, since insights only available from higher granular data are typically needed to derive actionable improvements. Understandably, differentiating for which processes higher granularity is needed is not always straightforward. Thus, this paper presents a guideline concept to support manufacturers in determining the required data and data granularity to identify improvement potential, while also assessing the ecological impact of their production. Firstly, energy consumption (electricity and heat), water and material use are determined to be the most relevant measures. Secondly, a structured categorization of production steps is created. As a starting point, selected metal working processes according to DIN 8580 are incorporated into the guideline. Thirdly, the data/device hierarchy from the Reference Architecture Model Industry 4.0 (RAMI 4.0) is used to define three data granularity levels. Finally, a structured approach for determining ideal data granularity for each production step is developed and applied to build an initial version of the guideline. An evaluation shows that the guideline recommends appropriate data granularities, promising to be a useful tool for manufacturers wanting to improve their ecological impact. Keywords Data granularity Environmental impact Production

8 February 2022 in Industrie 4.0 Management
Verified
Journal Article
Closed access
BGBIBA − Bremer Institut für Produktion und Logistik GmbH
BIBA − Bremer Institut für Produktion und Logistik GmbH
DBDennis Bode
Dennis Bode
Henry Ekwaro-OsireHenry Ekwaro-Osire
Henry Ekwaro-Osire
KTKlaus-Dieter Thoben
Klaus-Dieter Thoben
UBUniversität Bremen/Biba
Universität Bremen/Biba
Energieeffizienz gewinnt eine immer stärkere Bedeutung in allen Branchen des produzierenden Gewerbes. Mit Technologien aus den Bereichen Digitalisierung und Künstlicher Intelligenz (KI) können innovative Lösungen für eine energieeffizientere Herstellung beispielsweise durch eine KI-gestützte Prozessführung einen entscheidenden Beitrag zur Nachhaltigkeit und zur Wettbewerbsfähigkeit eines Unternehmens leisten. Die Anforderungen an den Einsatz und die reibungslose Integration von KI in Unternehmensprozesse sind komplex und vielfältig. Verschiedene Beteiligte müssen bei der Erfassung der Anforderungen integriert und unterschiedliche Technologieaspekte und Geschäftsfelder beachtet werden. Gegenstand dieses Beitrags ist ein Vorgehen zur Erfassung dieser Anforderungen bei der Durchführung innovativer Digitalisierungs- und KI-Projekte zur Steigerung der Energieeffizienz im produzierenden Gewerbe.