Lojistik depolarda sipariş toplama ve konumlandırmaya yönelik yenilikçi bir yaklaşım
Date
2023
Authors
Şahinaslan, Önder
Karataş, Ceyhun
Şahinaslan, Ender
Journal Title
Journal ISSN
Volume Title
Publisher
Sanayi ve Teknoloji Bakanlığı
Abstract
Amaç: Uluslararası bir lojistik deponun verimliliğini artırmaya katkı sağlamak için depo dolaşım mesafesini kısaltacak yenilikçi ve etkin bir konumlandırma yönteminin geliştirilmesi amaçlanmıştır.
Yöntem: 1.239.545 adet ham depo verisi uzman ekiplerin desteği ile analiz edilmiş ve yorumlanmıştır. Makine öğrenimi algoritmalarının depo konumlarının belirlenmesinde kullanılabilirliği araştırılmıştır. Dolaşım hızı ve anahtar değer hesaplamasına dayalı olarak alternatif bir konumlandırma çözümü geliştirilmiştir. Statik olarak beş farklı bölgeye ayrılan sahada uygulamalı testler yapılmıştır. Önerilen yöntemin etkinliği, bilinen konumlandırma yöntemleriyle karşılaştırılmıştır.
Bulgular: Öğrenme algoritmalarının başarı oranları (%54-%64) uzman ekipler tarafından yeterli bulunmamıştır. Geliştirilen çözümde ürünleri doğru yere yerleştirme başarı oranı %90,93 olmuştur. Bir aylık gözlem sonucunda depo giriş çıkış işlemlerinde kat edilen mesafe yaklaşık 880 km kısalmış, depo doluluk oranı %54,07'den %55,68'e yükselmiştir. Elde edilen sonuçlar önerilen yöntemin etkinliğini göstermektedir.
Özgünlük: Depo yerleşim yüzdeleri ve dolaşım mesafelerinde önemli kazanımlar elde edilmiştir. Bilinen diğer yöntemlere göre daha etkili ve yenilikçi bir yaklaşım sunmaktadır. Bölge sınırı olmayan dinamik, verimli ve başarılı yapısıyla farklı depolarda uygulanabilir özgünlüktedir. Gerçek depo verilerine ve uzman görüşlerine dayalı olarak oluşturulması literatüre eşsiz bir katkı sağlamaktadır.
Purpose: In order to contribute to increasing the efficiency of an international logistics warehouse, it is aimed to develop an innovative and effective positioning method that will shorten the warehouse circulation distance. Methodology: 1,239,545 raw warehouse data have been analyzed and interpreted with the support of expert teams. The usability of machine learning algorithms in determining warehouse locations has been investigated. An alternative positioning solution has been developed, based on circulation rate and key value calculation. Practical tests were carried out in the field, which is statically divided into five different regions. The effectiveness of the proposed method is compared with known positioning methods. Findings: The success rates (54%-64%) of the learning algorithms were not found sufficient by the expert teams. In the solution developed, the success rate of placing the products in the right place was 90.93%. As a result of a one-month observation, the distance covered in warehouse entry and exit operations has been shortened by approximately 880 km, and the warehouse occupancy rate has increased from 54.07% to 55.68%. Obtained results show the efficiency of the proposed method. Originality: Significant gains have been achieved in warehouse settlement percentages and circulation distances. It offers a more effective and innovative approach than other known methods. It is uniquely applicable in different warehouses with its dynamic and efficient structure that has no regional borders. Its creation based on real warehouse data and expert opinions makes a unique contribution to the literature.
Purpose: In order to contribute to increasing the efficiency of an international logistics warehouse, it is aimed to develop an innovative and effective positioning method that will shorten the warehouse circulation distance. Methodology: 1,239,545 raw warehouse data have been analyzed and interpreted with the support of expert teams. The usability of machine learning algorithms in determining warehouse locations has been investigated. An alternative positioning solution has been developed, based on circulation rate and key value calculation. Practical tests were carried out in the field, which is statically divided into five different regions. The effectiveness of the proposed method is compared with known positioning methods. Findings: The success rates (54%-64%) of the learning algorithms were not found sufficient by the expert teams. In the solution developed, the success rate of placing the products in the right place was 90.93%. As a result of a one-month observation, the distance covered in warehouse entry and exit operations has been shortened by approximately 880 km, and the warehouse occupancy rate has increased from 54.07% to 55.68%. Obtained results show the efficiency of the proposed method. Originality: Significant gains have been achieved in warehouse settlement percentages and circulation distances. It offers a more effective and innovative approach than other known methods. It is uniquely applicable in different warehouses with its dynamic and efficient structure that has no regional borders. Its creation based on real warehouse data and expert opinions makes a unique contribution to the literature.
Description
Keywords
Lojistik depo yönetimi , Etkili konumlandırma , Teknoloji ve yenilik , Verimlilik , Logistics warehouse management , Effective positioning , Technology and innovation , Productivity
Citation
Şahinaslan, Ö., Karatas, C. & Şahinaslan, E. (2023). Lojistik depolarda sipariş toplama ve konumlandırmaya yönelik yenilikçi bir yaklaşım. Verimlilik Dergisi, 57(3), 491-512. https://doi.org/10.51551/verimlilik.1188635.
