Integrating off-site and on-site panelized construction schedules using fleet dispatching

Highlights

• A case study is conducted for residential projects using panelized construction.

• A fleet dispatching framework is proposed to enhance overall operational efficiency.

• Schedules linking a factory to sites can be optimized through logistics operations.

• It achieves 90% of on-time deliveries and 40-min early completion of daily works.

• Schedule optimization reduces operational delays and under-utilization of equipment.

Abstract

In off-site construction, finding good synchronization between the factory and the construction site is critically important. However, in practice, on- and off-site operations are typically scheduled separately, without considering their performance at the project level. For the planning of logistics operations that bridge a factory and sites, this paper proposes a framework to optimize the truck-dispatching schedule using a discrete-event simulation model built based on real fleet operational data (e.g., GPS). Independently pre-planned schedules of factory and site operations are used as inputs, and historical data from transportation equipment are leveraged to generate the dispatching schedule. According to actual panelized residential projects, the proposed framework resulted in approximately 90% of deliveries being made on time, and daily activities were completed on average 40 min earlier than the historical time. Thus, the optimization of on- and off-site schedules may reduce the delays and under-utilization of equipment.

Introduction

In recent decades, off-site construction methods (e.g., modular or panelized) have been increasingly pursued to improve operational efficiency and resolve labor shortage and low productivity issues [1]. In particular, panelized construction has attracted considerable attention owing to its advantages over the modular construction method with regard to design flexibility and transportation requirements [2]. Unlike three-dimensional (3D) modular construction, that requires relatively higher capacity trailers and cranes, panelized construction can be applied various designs, while two-dimensional (2D) prefabricated panels can still be easily assembled on a site with minimal transportation demand (i.e., smaller equipment capacity). Therefore, panelized construction has emerged as a promising construction method in the residential construction sector. Previous studies have reported its benefits in terms of project delivery time, construction cost, site safety, and ecological footprint [3]. However, the actual adoption of this method has been slow, and most residential construction companies in major markets still use a traditional “stick-built” approach [4]. Recent statistics from the US Census Bureau show that less than 2% of new single-family homes are built using an off-site construction approach [5]. To effectively implement panelized construction and realize its potential benefits, an in-depth understanding of and sound decision-making with respect to the panelized construction rooted in actual operational data is essential. However, this has not been readily available to the construction industry because of the lack of a systematic, data-driven decision-making approach.

One of the main reasons for the slow adoption of panelized construction is the lack of large-scale operational data and advanced management tools [6], and logistics operations are particularly challenging in this respect. While panelized residential construction can be broadly understood in terms of the two major operations of panel pre-fabrication (factory operations) and on-site assembly (site operations), effective logistics management is the key to achieving overall operational efficiency. For example, fleets of trucks and trailers are utilized to transport prefabricated panels from factories to multiple construction sites. Without a precisely coordinated flow of completed panels from the factory to multiple construction sites, both factories and sites may be adversely affected in terms of productivity. At the factory, if completed panels are not picked up from the panel production line on time, there is a risk of production stoppage due to limited storage space. Although a factory has buffer areas for storing excess completed panels, these areas should ideally be minimized to ensure space efficiency in the plant. At the construction site, a delay in the delivery of the completed panels results in idle time for the crane, thereby increasing costs. Moreover, similar to a factory where storage space is limited, construction sites often have limited storage areas for parking delivery trailers, which may become necessary if plant and site operations are not well coordinated. In addition to the spatial constraints at both the factory and the site, owing to finite logistics resources (e.g., a trailer), delivery trailers need to be returned to the factory as soon as possible following the delivery of completed panels to a site so that they are available for the next load of completed panels queued for delivery. Thus, the operational schedules of the factory and construction sites should be developed synchronously to achieve high productivity and minimize project delays. Moreover, the capacities of trailers and trucks must also be considered during the planning process.

For the present study, we consider the case of an actual panelized construction company based in Edmonton, Canada. Operators at this facility report that a logistics (e.g., truck dispatching) schedule is a critical element that could be introduced to harmonize factory and site operations. Despite its potential positive impact in this respect, the notion of a framework for coordinated logistics operation is still ill-understood, and logistics has received relatively little attention in off-site construction compared to factory and site operations. In the case of a collaborating partner, logistics operations are typically managed in an ad hoc manner (e.g., via phone calls and text messages between operators (e.g., loader, drivers, crane operator, etc.)). In addition, panelized construction companies place a disproportionate emphasis on the delivery of panels to sites on time to achieve high crane productivity (minimize idle time). As a result of this site-focused approach, the production line often needs to be held until empty delivery trailers become available. In this context, a more thorough understanding of the complex interrelationships among off-site prefabrication, logistics, and site operations is required to accurately compare panelized construction with the traditional stick-built method and gain a better understanding of and capitalize on the benefits of panelized construction in terms of operational efficiency.

To generate a schedule for logistics operations that optimally harmonizes both plant and site operations in panelized residential construction, this paper proposes a framework that allows operators and managers to gain a better understanding of operational capacity while improving operational control. The proposed framework optimizes the logistics schedule using a discrete-event simulation (DES) model and by leveraging operational data (e.g., activity, location, and time) in cloud and local databases (DBs). To efficiently collect data from both the prefabrication facility and the site, (1) a web application based on a cloud DB is developed to collect operational data from multiple operators using a quick response (QR) code system to minimize manual input; (2) both the cloud and local databases are designed to collect and process the data efficiently, and to generate the schedule; and (3) a DES model with a heuristic optimization approach is applied to generate the optimum fleet operation schedule. To examine and validate the proposed framework, a case study is conducted at a panelized residential construction company in Edmonton, Canada. During the panelized construction operations, the framework could be applied to improve the overall project operations by optimizing the factory and site schedules through simulated logistic dispatching sequences. The optimized dispatching schedules, which indicate when and where to deliver products among various operational alternatives, may support operation managers in improving their decision-making processes. The generated dispatching sequences can also be periodically processed to continually optimize the operation schedule based on the current situation.