Improving bicycle sharing operations: A multi-criteria decision-making approach

Highlights

• Considering attribute relationships and weights, a supermatrix is proposed.

• Economic cost is crucial to the operation and management of bicycle-sharing.

• Prioritize connectivity between bicycle-sharing and destinations for flexibility.

• Preferential rent is an effective method to improve the service level.

Abstract

Bicycle sharing is quickly becoming popular because it is environmentally friendly and offers convenience to customers. Determining how to improve the service level and customer satisfaction are the keys to a successful bicycle sharing business in today’s fiercely competitive market and are also the keys to achieving a sustainable operation. In this paper, following a hierarchical structure, five main factors related to the bicycle sharing operations are identified: (1) convenience and flexibility, (2) operation service, (3) economic cost, (4) design and layout and (5) management specification. Eighteen sub factors are then derived. To determine the significant factors affecting bicycle sharing operations, the Multi-Criteria Decision-Making (MCDM) approach based on Interval Type-2 Fuzzy Sets (IT2FSs), Decision Making Trial and Evaluation Laboratory Model (DEMATEL) and the Analytic Network Process (ANP) are proposed. First, IT2FSs and fuzzy DEMATEL are integrated to analyze the complex relationships among the various factors. DEMATEL is then integrated with ANP to rank the importance of these factors. Results are compared with those of other state-of-the-art MCDM methods. It is concluded that convenience and flexibility, rationality of the layout, and economic cost are the three most important factors impacting the bicycle sharing business model. Based on the findings, theoretical guidance for the development of bicycle sharing operations is provided.

Introduction

As a new product of the Internet economy, bicycle sharing provides users with a more convenient mode of transportation. Compared with motor vehicles, bicycle sharing also reduces air pollution, carbon emissions and waste of non-renewable energy and promotes the rapid development of green economy. Importantly, more and more people are willing to avoid traditional public transportation (bus and rail) and walk or ride their bikes due to the influence of COVID-19, which will bring a new upsurge of bicycle sharing, a breakthrough innovation for a city’s slow transportation system.¹ It has successfully transformed the bicycle travel mode into a new urban fashion and has quickly emerged as one of the important symbols of the sharing economy. According to statistics from the Ministry of Transport in China, there were nearly 70 bicycle sharing companies across the country as of the end of July 2017. More than 16 million vehicles had been put into service and up to 1.5 billion passengers have been served (Kadri et al., 2016; Ministry of Transport in China, 2017). By August 2019, 19.5 million bicycles have been rented on the Internet in China, covering 360 cities across the country, with more than 300 million registered users, making it the third largest public transportation tool in China, behind only bus and rail transit.² In addition, by April 2020, affected by the epidemic, the average daily cycling volume of Meituan bikes in China has increased by 410% compared with that in February. According to the data of green orange bikes, by March 2020, the daily cycling volume of shared bikes has recovered more than 80% of that of the same period in 2019, and the growth of first and second-tier cities is even more obvious.³ At the same time, bicycle sharing is also a concrete manifestation of “Internet + Travel”, which extends the accessibility of public transport systems to the customers’ final destinations and helps solve the “last mile” problem in public transport modes. In the bicycle sharing system (as shown in Fig. 1), there are three flows. The material flow reflects the movement of the bicycles sharing among the operators and customers. The flow of information is generated throughout the bicycle sharing system, which is critical for the operator. It can help the operator to obtain basic information about the market and its customers. In the financial flow, the operator is the main profit organization; it invests funds to maintain the operation and retains a large proportion of the profits.

Due to its environmental protection and convenience, bicycle sharing has not only become a social hotspot, but has also attracted widespread attention from scholars at home and abroad. Currently, the topics of site selection and optimal path planning for sharing bicycles have become significant research directions. However, in reality, issues such as the handling problems of abandoned bicycles, shared bicycles affecting urban traffic and so on are widely discussed, which indicates that the operation of sharing bicycles is encountering difficulties, thus affecting its sustainable development. Therefore, identifying the important factors affecting bicycle sharing operations and proposing potential improvement measures is becoming critical. The main purpose of this paper is to select the most significant factors from the many factors affecting the operation level of bicycle sharing. There are two questions that need to be explained in the research methodologies, which are as follows: Q1: Why are the Decision Making Trial and Evaluation Laboratory Model (DEMATEL) and the Analytic Network Process (ANP) chosen in many Multi-Criteria Decision-Making (MCDM) approaches based on Interval Type-2 Fuzzy Sets (IT2FSs)? Q2: Why are DEMATEL & IT2FSs and DEMATEL & ANP integrated?

In terms of Q1, DEMATEL is commonly used to analyze the relationship between criteria, and ANP is often used to determine the weights of the criteria. Typically, there exists no completely independent relationship between the criteria; therefore, the relationships between the criteria and the importance of the criteria both need to be considered in MCDM. Convenience and flexibility, Operation service, Economic cost, Design and layout, and Management specification are five key factors that affect the efficiency of sharing bicycles. More importantly, each of these key factors has sub-criteria that are critical to improving the operation and management of a sharing bicycle system. The combination of ANP and DEMATEL methods can not only help decision makers to analyze the importance of related criteria, but also to analyze the correlation between factors. For Q2, On the one hand, experts should be invited to evaluate these criteria and an accurate evaluation language can reduce decision errors. The IT2FSs can more accurately reflect the decision-maker’s attitude than the 1–9 scale method and triangular fuzzy numbers (TFN). Therefore, it is reasonable to combine IT2FSs with DEMTAEL to investigate the relationship between the criteria. On the other hand, identifying the key factors affecting the operational efficiency of sharing bicycles and analyzing the correlations between these factors is essential for proposing effective operating strategies. ANP is mostly used to study the weights of the criteria with the associated relationships. Therefore, the analysis of association degree between criteria is not only the basis for the weights investigation of ANP, but also an important basis to establish a super decision matrix.

This paper is structured as follows: section 2 is the literature review, which is related to the bicycle sharing operations and MCDM approaches. In section 3, the research methods used in this paper are introduced including IT2FSs, the Decision Making Trial and Evaluation Laboratory Model (DEMATEL) and ANP. In section 4, five major factors that affect the service level of bicycle sharing operations and the corresponding 18 sub factors are analyzed. A case is studied in section 5. Section 6 summarizes the research results and some feasible comments are provided to improve the service level and customer satisfaction of the bicycle sharing operators based on the different types of customers.