Premium
Using horizontal cooperation concepts in integrated routing and facility‐location decisions
Author(s) -
QuinteroAraujo Carlos L.,
Gruler Aljoscha,
Juan Angel A.,
Faulin Javier
Publication year - 2019
Publication title -
international transactions in operational research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.032
H-Index - 52
eISSN - 1475-3995
pISSN - 0969-6016
DOI - 10.1111/itor.12479
Subject(s) - supply chain , vehicle routing problem , flexibility (engineering) , benchmark (surveying) , computer science , routing (electronic design automation) , operations research , business , economics , marketing , computer network , management , geodesy , engineering , geography
In a global and competitive economy, efficient supply networks are essential for modern enterprises. Horizontal cooperation (HC) concepts represent a promising strategy to increase the performance of supply chains. HC is based on sharing resources and making joint decisions among different agents at the same level of the supply chain. This paper analyzes different cooperation scenarios concerning integrated routing and facility‐location decisions in road transportation: (a) a noncooperative scenario in which all decisions are individually taken (each enterprise addresses its own vehicle routing problem [VRP]); (b) a semicooperative scenario in which route‐planning decisions are jointly taken (facilities and fleets are shared and enterprises face a joint multidepot VRP); and (c) a fully cooperative scenario in which route‐planning and facility‐location decisions are jointly taken (also customers are shared, and thus enterprises face a general location routing problem). Our analysis explores how this increasing level of HC leads to a higher flexibility and, therefore, to a lower total distribution cost. A hybrid metaheuristic algorithm, combining biased randomization with a variable neighborhood search framework, is proposed to solve each scenario. This allows us to quantify the differences among these scenarios, both in terms of monetary and environmental costs. Our solving approach is tested on a range of benchmark instances, outperforming previously reported results.