An Object-Oriented Algorithm for Flexible Construction Scheduling with Automated Task Splitting

Read full text
Authors: José Antonio González-Fajardo, Jesús Nicolás Zaragoza-Grifé, Sergio Omar Álvarez Romero
Cited by: 0
References: 13

Abstract

The Precedence Diagram Method (PDM) is widely used for modeling large and complex construction projects characterized by non-linear execution logic, overlapping activities, and multiple precedence constraints. Despite involving standardized and recurrent operations (e.g., formwork, reinforcement, concreting), construction schedules often exhibit highly complex logic that challenges traditional Critical Path Method (CPM) analysis. In particular, PDM networks with multiple relationships complicate critical path identification and float determination, while current commercial scheduling tools impose restrictive logic and limited support for activity splitting. This study proposes a novel object-oriented algorithm and a supporting software tool that integrates automated multi-activity splitting, unrestricted handling of precedence types, and segment-level criticality identification within a unified scheduling framework. The proposed approach transforms complex PDM networks into equivalent Finish-to-Start-based structures solvable using conventional forward and backward pass algorithms, while strictly preserving the original scheduling logic. Results demonstrate that the proposed method improves analytical transparency, scheduling flexibility, and precision in critical path and float analysis, including the identification of partially critical activity segments. The proposed framework provides project managers with a practical and scalable solution to optimize productivity and reduce project duration in complex construction scheduling scenarios.

Fields of Study & Subjects

Resource-Constrained Project Scheduling Construction Project Management and Performance BIM and Construction Integration

Publication details

Journal / source
Journal of Building Construction and Planning Research
Volume
14
Issue
01
Pages
21–40
Publication date
1 January 2026