Application of Machine Learning on Transport Spot Rate Prediction In the Recycling Industry
Research output: Contributions to collected editions/works › Article in conference proceedings › Research › peer-review
Standard
Conference on Production Systems and Logistics: International Conference, CPSL 2022, hosted at the University of British Columbia in Vancouver, Canada, 17th May 2022 – 20th May 2022, Proceedings. ed. / David Herberger; Marco Hübner. Hannover: publish-Ing., 2022. p. 554-563 (Proceedings of the Conference on Production Systems and Logistics; Vol. 3).
Research output: Contributions to collected editions/works › Article in conference proceedings › Research › peer-review
Harvard
APA
Vancouver
Bibtex
}
RIS
TY - CHAP
T1 - Application of Machine Learning on Transport Spot Rate Prediction In the Recycling Industry
AU - Green, Thorben
AU - Rokoss, Alexander
AU - Kramer, Kathrin
AU - Schmidt, Matthias
N1 - Funding Information: The study was carried out in the framework of the research project Ä.,-Werkstatt: Künstliche Intelligenz in 3URGXNWLRQVXQWHUQHKPHQ .,:H ³. It is supported by the European Union and the State of Lower Saxony within the EFRE program. Publisher Copyright: © 2022, Publish-Ing in cooperation with TIB - Leibniz Information Centre for Science and Technology University Library. All rights reserved.
PY - 2022
Y1 - 2022
N2 - The transport spot rate in trucking logistics is an important factor for market participants in the recycling industry. Knowledge about the current spot rate is essential for operational decision-making in price negotiations between brokers and shippers. Due to the characteristics and dynamics of the industry, this task is particularly challenging. So far, businesses mainly rely on traditional calculation methods combined with their own expertise in price negotiations. The growing amount of existing business and market data may enable companies to take advantage of data-driven decision processes. However, the resulting volume of data and required effort for analysis do not match the fast pace of daily business. To improve current forecasting practices, this paper conducts a comparative study of machine learning (ML) approaches for shipment-specific spot rate prediction. For this, the paper builds on the experience and database of a small broker in the recycling industry in Northern Germany and complements it with external market information. The study shows the ability of ML to internalize underlying patterns between spot rates and market data. During the use case the CRISP-DM framework is followed to select the most appropriate features and train multiple ML algorithms. Several metrics are applied to determine the most accurate model for spot rate prediction. Results indicate that especially the ML-algorithm Random Forest shows considerable potential to provide brokers in the recycling industry with more reliable spot rate assumptions. Therefore, future implementation of ML approaches in the industry may open up new and beneficial business opportunities. The study paths the way for further research on the predictive potential of ML for prices in transportation with extended and diversified data sets.
AB - The transport spot rate in trucking logistics is an important factor for market participants in the recycling industry. Knowledge about the current spot rate is essential for operational decision-making in price negotiations between brokers and shippers. Due to the characteristics and dynamics of the industry, this task is particularly challenging. So far, businesses mainly rely on traditional calculation methods combined with their own expertise in price negotiations. The growing amount of existing business and market data may enable companies to take advantage of data-driven decision processes. However, the resulting volume of data and required effort for analysis do not match the fast pace of daily business. To improve current forecasting practices, this paper conducts a comparative study of machine learning (ML) approaches for shipment-specific spot rate prediction. For this, the paper builds on the experience and database of a small broker in the recycling industry in Northern Germany and complements it with external market information. The study shows the ability of ML to internalize underlying patterns between spot rates and market data. During the use case the CRISP-DM framework is followed to select the most appropriate features and train multiple ML algorithms. Several metrics are applied to determine the most accurate model for spot rate prediction. Results indicate that especially the ML-algorithm Random Forest shows considerable potential to provide brokers in the recycling industry with more reliable spot rate assumptions. Therefore, future implementation of ML approaches in the industry may open up new and beneficial business opportunities. The study paths the way for further research on the predictive potential of ML for prices in transportation with extended and diversified data sets.
KW - Engineering
KW - Machine learning
KW - Business informatics
KW - transport spot rate
KW - price prediction
KW - Reverse Logistics
KW - Recycling
UR - https://cpsl-conference.com/wp-content/uploads/2022/06/Proceedings-of-the-Conference-on-Production-Systems-and-Logistics-CPSL-2022.pdf
UR - http://www.scopus.com/inward/record.url?scp=85164377283&partnerID=8YFLogxK
U2 - 10.15488/12151
DO - 10.15488/12151
M3 - Article in conference proceedings
T3 - Proceedings of the Conference on Production Systems and Logistics
SP - 554
EP - 563
BT - Conference on Production Systems and Logistics
A2 - Herberger, David
A2 - Hübner, Marco
PB - publish-Ing.
CY - Hannover
T2 - 3rd Conference on Production Systems and Logistics - CPSL 2022
Y2 - 17 May 2022 through 20 May 2022
ER -