. Metzger, A., Franke, J., & Jansen, T. vom Brocke, J., Mendling, J., & Rosemann, M., editors. Ensemble Deep Learning for Proactive Terminal Process Management at the Port of Duisburg ``duisport'', pages 153–164. Springer Berlin Heidelberg, Berlin, Heidelberg, 2021. ![Ensemble Deep Learning for Proactive Terminal Process Management at the Port of Duisburg ``duisport'' [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex (a)Situation faced: The case presented here is located at duisport, the world's largest inland container port. Duisport is situated at the center of Germany's largest metropolitan region, with close to 10 million inhabitants. A multitude of roads, tracks, and waterways serve as entry and exit points for containers to and from duisport, and this transport infrastructure is shared with the metropolitan region. Thus, any increase in container volumes due to the overall growth in freight transport cannot be captured by growing in terms of space but requires improvement in the terminal's productivity.(b)Action taken: We employed advanced data analytics to provide decision support for terminal operators and facilitate proactive management of the terminal's processes. Using ensembles of deep learning models, we prototypically developed the Terminal Productivity Cockpit (TPC). The TPC predicts delays in the execution of a running process and provides operators with decision support regarding whether to intervene by adapting the process. The case employed more than 30 million data entries from eight data sources, such as cranes and towing vehicles.(c)Results achieved: We estimated that, by using the TPC, terminal operators may increase the rate of successful process outcomes by 4.7%, leading to estimated cost savings of up to 22.5%.(d)Lessons learned: Big data and deep learning offer opportunities for management and innovation of transport processes. We observed that ensemble deep learning contributes to the efficient engineering of data-driven applications because it works well without extensive feature engineering and hyper-parameter tuning. Still, the effectiveness of data analytics solutions depends to a large degree on the quality of data. In this case, understanding, integrating, and cleansing of data consumed around 80% of the overall development time and resources.
@Inbook{Metzger2021,
author="Metzger, Andreas
and Franke, Johannes
and Jansen, Thomas",
editor="vom Brocke, Jan
and Mendling, Jan
and Rosemann, Michael",
title="Ensemble Deep Learning for Proactive Terminal Process Management at the Port of Duisburg ``duisport''",
bookTitle="Business Process Management Cases Vol. 2: Digital Transformation - Strategy, Processes and Execution",
year="2021",
publisher="Springer Berlin Heidelberg",
address="Berlin, Heidelberg",
pages="153--164",
abstract="(a)Situation faced: The case presented here is located at duisport, the world's largest inland container port. Duisport is situated at the center of Germany's largest metropolitan region, with close to 10 million inhabitants. A multitude of roads, tracks, and waterways serve as entry and exit points for containers to and from duisport, and this transport infrastructure is shared with the metropolitan region. Thus, any increase in container volumes due to the overall growth in freight transport cannot be captured by growing in terms of space but requires improvement in the terminal's productivity.(b)Action taken: We employed advanced data analytics to provide decision support for terminal operators and facilitate proactive management of the terminal's processes. Using ensembles of deep learning models, we prototypically developed the Terminal Productivity Cockpit (TPC). The TPC predicts delays in the execution of a running process and provides operators with decision support regarding whether to intervene by adapting the process. The case employed more than 30 million data entries from eight data sources, such as cranes and towing vehicles.(c)Results achieved: We estimated that, by using the TPC, terminal operators may increase the rate of successful process outcomes by 4.7{\%}, leading to estimated cost savings of up to 22.5{\%}.(d)Lessons learned: Big data and deep learning offer opportunities for management and innovation of transport processes. We observed that ensemble deep learning contributes to the efficient engineering of data-driven applications because it works well without extensive feature engineering and hyper-parameter tuning. Still, the effectiveness of data analytics solutions depends to a large degree on the quality of data. In this case, understanding, integrating, and cleansing of data consumed around 80{\%} of the overall development time and resources.",
isbn="978-3-662-63047-1",
doi="10.1007/978-3-662-63047-1_12",
url="https://doi.org/10.1007/978-3-662-63047-1_12"
}
Downloads: 0
{"_id":"X3AF9GrRA7hiaH4vB","bibbaseid":"metzger-franke-jansen-ensembledeeplearningforproactiveterminalprocessmanagementattheportofduisburgduisport-2021","author_short":["Metzger, A.","Franke, J.","Jansen, T."],"bibdata":{"bibtype":"inbook","type":"inbook","author":[{"propositions":[],"lastnames":["Metzger"],"firstnames":["Andreas"],"suffixes":[]},{"propositions":[],"lastnames":["Franke"],"firstnames":["Johannes"],"suffixes":[]},{"propositions":[],"lastnames":["Jansen"],"firstnames":["Thomas"],"suffixes":[]}],"editor":[{"propositions":["vom"],"lastnames":["Brocke"],"firstnames":["Jan"],"suffixes":[]},{"propositions":[],"lastnames":["Mendling"],"firstnames":["Jan"],"suffixes":[]},{"propositions":[],"lastnames":["Rosemann"],"firstnames":["Michael"],"suffixes":[]}],"title":"Ensemble Deep Learning for Proactive Terminal Process Management at the Port of Duisburg ``duisport''","booktitle":"Business Process Management Cases Vol. 2: Digital Transformation - Strategy, Processes and Execution","year":"2021","publisher":"Springer Berlin Heidelberg","address":"Berlin, Heidelberg","pages":"153–164","abstract":"(a)Situation faced: The case presented here is located at duisport, the world's largest inland container port. Duisport is situated at the center of Germany's largest metropolitan region, with close to 10 million inhabitants. A multitude of roads, tracks, and waterways serve as entry and exit points for containers to and from duisport, and this transport infrastructure is shared with the metropolitan region. Thus, any increase in container volumes due to the overall growth in freight transport cannot be captured by growing in terms of space but requires improvement in the terminal's productivity.(b)Action taken: We employed advanced data analytics to provide decision support for terminal operators and facilitate proactive management of the terminal's processes. Using ensembles of deep learning models, we prototypically developed the Terminal Productivity Cockpit (TPC). The TPC predicts delays in the execution of a running process and provides operators with decision support regarding whether to intervene by adapting the process. The case employed more than 30 million data entries from eight data sources, such as cranes and towing vehicles.(c)Results achieved: We estimated that, by using the TPC, terminal operators may increase the rate of successful process outcomes by 4.7%, leading to estimated cost savings of up to 22.5%.(d)Lessons learned: Big data and deep learning offer opportunities for management and innovation of transport processes. We observed that ensemble deep learning contributes to the efficient engineering of data-driven applications because it works well without extensive feature engineering and hyper-parameter tuning. Still, the effectiveness of data analytics solutions depends to a large degree on the quality of data. In this case, understanding, integrating, and cleansing of data consumed around 80% of the overall development time and resources.","isbn":"978-3-662-63047-1","doi":"10.1007/978-3-662-63047-1_12","url":"https://doi.org/10.1007/978-3-662-63047-1_12","bibtex":"@Inbook{Metzger2021,\r\nauthor=\"Metzger, Andreas\r\nand Franke, Johannes\r\nand Jansen, Thomas\",\r\neditor=\"vom Brocke, Jan\r\nand Mendling, Jan\r\nand Rosemann, Michael\",\r\ntitle=\"Ensemble Deep Learning for Proactive Terminal Process Management at the Port of Duisburg ``duisport''\",\r\nbookTitle=\"Business Process Management Cases Vol. 2: Digital Transformation - Strategy, Processes and Execution\",\r\nyear=\"2021\",\r\npublisher=\"Springer Berlin Heidelberg\",\r\naddress=\"Berlin, Heidelberg\",\r\npages=\"153--164\",\r\nabstract=\"(a)Situation faced: The case presented here is located at duisport, the world's largest inland container port. Duisport is situated at the center of Germany's largest metropolitan region, with close to 10 million inhabitants. A multitude of roads, tracks, and waterways serve as entry and exit points for containers to and from duisport, and this transport infrastructure is shared with the metropolitan region. Thus, any increase in container volumes due to the overall growth in freight transport cannot be captured by growing in terms of space but requires improvement in the terminal's productivity.(b)Action taken: We employed advanced data analytics to provide decision support for terminal operators and facilitate proactive management of the terminal's processes. Using ensembles of deep learning models, we prototypically developed the Terminal Productivity Cockpit (TPC). The TPC predicts delays in the execution of a running process and provides operators with decision support regarding whether to intervene by adapting the process. The case employed more than 30 million data entries from eight data sources, such as cranes and towing vehicles.(c)Results achieved: We estimated that, by using the TPC, terminal operators may increase the rate of successful process outcomes by 4.7{\\%}, leading to estimated cost savings of up to 22.5{\\%}.(d)Lessons learned: Big data and deep learning offer opportunities for management and innovation of transport processes. We observed that ensemble deep learning contributes to the efficient engineering of data-driven applications because it works well without extensive feature engineering and hyper-parameter tuning. Still, the effectiveness of data analytics solutions depends to a large degree on the quality of data. In this case, understanding, integrating, and cleansing of data consumed around 80{\\%} of the overall development time and resources.\",\r\nisbn=\"978-3-662-63047-1\",\r\ndoi=\"10.1007/978-3-662-63047-1_12\",\r\nurl=\"https://doi.org/10.1007/978-3-662-63047-1_12\"\r\n}\r\n\r\n","author_short":["Metzger, A.","Franke, J.","Jansen, T."],"editor_short":["vom Brocke, J.","Mendling, J.","Rosemann, M."],"key":"Metzger2021","id":"Metzger2021","bibbaseid":"metzger-franke-jansen-ensembledeeplearningforproactiveterminalprocessmanagementattheportofduisburgduisport-2021","role":"author","urls":{"Paper":"https://doi.org/10.1007/978-3-662-63047-1_12"},"metadata":{"authorlinks":{}}},"bibtype":"inbook","biburl":"https://paluno.uni-due.de/fileadmin/sse/user_upload/SSE_Publikationen.bib","dataSources":["ywizExvTQ8JG7S3Wa"],"keywords":[],"search_terms":["ensemble","deep","learning","proactive","terminal","process","management","port","duisburg","duisport","metzger","franke","jansen"],"title":"Ensemble Deep Learning for Proactive Terminal Process Management at the Port of Duisburg ``duisport''","year":2021}