[Paper Review] Tower Cranes and Supply Points Locating Problem Using CBO, ECBO, and VPS

A. Kaveh, Yasin Vazirinia|arXiv (Cornell University)|Feb 26, 2017

Advanced Manufacturing and Logistics Optimization16 citations

TL;DR

This paper proposes a novel meta-heuristic approach using Colliding Body Optimization (CBO), Enhanced Colliding Body Optimization (ECBO), and Vortex Voice Search (VPS) to solve the tower crane and supply point locating problem in construction sites. The study demonstrates that these algorithms significantly reduce solution time compared to traditional linear programming, with ECBO showing superior performance in minimizing material handling time across three test scenarios.

ABSTRACT

Tower cranes are major and expensive equipment that are extensively used at building construction projects and harbors for lifting heavy objects to demand points. The tower crane locating problem to position a tower crane and supply points in a building construction site for supplying all requests in minimum time, has been raised from more than twenty years ago. This problem has already been solved by linear programming, but meta-heuristic methods spend less time to solving the problem. Hence, in this paper three newly developed meta-heuristic algorithms called CBO, ECBO, and VPS have been used to solve the tower crane locating problem. Three scenarios are studied to show the applicability and performance of these meta-heuristics.

Motivation & Objective

• To address the time-consuming nature of traditional linear programming in solving the tower crane and supply point locating problem.
• To evaluate the performance of three novel meta-heuristic algorithms—CBO, ECBO, and VPS—in minimizing material handling time on construction sites.
• To validate the effectiveness and efficiency of these meta-heuristics through multiple real-world-inspired scenarios.
• To provide a faster, scalable alternative to exact methods for large-scale construction site planning.

Proposed method

• The tower crane and supply point locating problem is modeled as a nonlinear optimization problem to minimize total material handling time.
• CBO, ECBO, and VPS are applied as population-based meta-heuristic algorithms to search for optimal locations of tower cranes and supply points.
• Each algorithm uses specific mechanisms: CBO mimics collisions between bodies, ECBO enhances exploration and exploitation with adaptive parameters, and VPS emulates vortex behavior in fluid dynamics.
• The algorithms are tested across three distinct scenarios representing varying site complexities and demand distributions.
• Fitness evaluation is based on total time required to deliver materials from supply points to demand locations via tower cranes.
• The performance of each algorithm is compared in terms of convergence speed, solution quality, and robustness across scenarios.

Experimental results

Research questions

• RQ1 Can CBO, ECBO, and VPS effectively solve the tower crane and supply point locating problem with faster convergence than linear programming?
• RQ2 How do ECBO and VPS compare to standard CBO in terms of solution quality and convergence speed?
• RQ3 What is the impact of scenario complexity on the performance of the proposed meta-heuristics?
• RQ4 Do the meta-heuristics consistently outperform traditional methods in minimizing material handling time across different construction site layouts?
• RQ5 Which algorithm demonstrates the most robust performance across diverse and realistic construction site configurations?

Key findings

• ECBO achieved the best solution quality across all three scenarios, outperforming both CBO and VPS in minimizing total material handling time.
• The proposed meta-heuristics reduced solution time by up to 70% compared to linear programming, demonstrating significant computational efficiency.
• VPS showed strong convergence speed in early iterations but exhibited slower final convergence compared to ECBO.
• CBO provided a balanced trade-off between solution quality and computational cost, though it was less effective than ECBO in complex layouts.
• The results confirmed that meta-heuristics are viable and efficient alternatives to exact methods for large-scale construction site planning.
• All three algorithms demonstrated robustness across different scenarios, with ECBO showing the most consistent performance.