Automation in Construction - Article - Learning and critiquing pairwise activity relationships for schedule quality control via deep learning-based natural language processing

Oct 22, 2026

Construction schedule mistakes often lead to delays and costly liquidated damages, and industry guidelines require planners to verify schedule logic. Logic checking is manual and knowledge-intensive because existing tools do not reliably validate missing or incorrect logical dependencies. This paper proposes an automatic machine-learning approach that learns scheduling domain knowledge from records and validates input schedules, achieving an F1 score of 88.3%. It also extends the learned knowledge to schedule unordered activities, discussing method, experiments, benefits, and limitations.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages24

Reading Time1 h 0 min

AccessFree to download

Introduction

Background and motivation

Related work on automation and metrics

Pairwise relationship critique

Method and learning from records

Experimental results

Discussion: benefits and limitations

Automation in Construction 134 (2022) 104036 Contents lists available at ScienceDirect Automation in Construction journal [homepage:](homepage: www.elsevier.com/locate/autcon)[ www.elsevier.com/locate/autcon](homepage: www.elsevier.com/locate/autcon) | Learning and critiquing pairwise activity relationships for schedule quality control via deep learning-based natural language processing
Fouad Amera, *, Julia Hockenmaier b, Mani Golparvar-Fardc
a Department of Civil and Environmental Engineering and Department of Computer Science, University of Illinois at Urbana-Champaign, 205 N. MathewsAve., Urbana, IL 61801, USA
b Computer Science, University of Illinois at Urbana-Champaign, 201 N. Goodwin Ave., Urbana, IL 61801, USA
c Civil and Environmental Engineering, Computer Science, and Technology Entrepreneurship, University of Illinois at Urbana-Champaign, 205 N. Mathews Ave., Urbana, IL 61801, USA | | | | --- | --- | --- | | A R T I C L E I N F O | | A B S T R A C T | | Keywords:
Construction planning Quality control
Data mining
Natural language processing Machine learning
Neural networks | | In construction, schedule mistakes causing delays beyond substantial completion dates cost contractors expensive liquidated damages. Hence, several industry guidelines, such as the DCMA's 14 point assessment, define schedule quality and offer systematic methods for ensuring it. These guidelines list “logic” as an essential control metric, and they require planners to ensure their schedules are free of missing or wrong logical dependencies. Checking the logic requires extensive construction domain knowledge, and planners perform it entirely manually as there are no available software solutions that support it. This paper offers a novel machine learning-based solution that learns construction scheduling domain knowledge from existing records completely automatically and applies it to validate the logic in input schedules achieving an F1 score of 88.3%. Furthermore, we tailor our method to use the learned knowledge to schedule a list of unordered activities. The details of the method, experimental results, benefits, and limitations are discussed. | 1. Introduction Construction planning is a knowledge-intensive task where planners accumulate data from different sources and in various formats, envision a process to realize the project, and communicate that vision to different project stakeholders in form of a project schedule. The sheer volume of data needed in form of a project scope –usually made available though quantity take-off– and the multitude of the dependencies among activities in forms of structural, functional, legal, resource, information constraints, make planning large projects a difficult undertaking. According to McKinsey & Company [1], poor planning, among other factors, causes 98% of projects to be delayed by 20 months and to incur 80% increase in cost. To that end, industry guidelines such as those developed by the Project Management Institute (PMI), the Defense Contract Management Agency (DCMA), the Government Accountability Office (GAO), and the National Defence Industrial Association- Integrated Program Management Division (NDIA-IPMD) [2–5] as well as academic research such as [6–8] stress the importance of scheduling logic in insuring high-quality schedules and offer procedures and checklists to validate logical dependencies. However, when schedule logic and its impact on the project duration is considered, it is easy to see how small human errors can lead to dire consequences. All too often, mistakes are made by having too many constraints, neglecting trade logic or stacking trades with too many start-to-start and finish-to-finish relationships. In cases where logic is missing, the schedule has a lot of float embedded in it, which usually results in delays going undetected until late in the project execution phase. The labor-intensive and experience-driven nature of planning can

FAQs

Why is schedule logic validation important in construction planning?

Schedule logic mistakes can cause project delays and financial penalties. Industry guidance emphasizes ensuring schedules have no missing or wrong logical dependencies.

What does the proposed paper contribute?

It presents a machine-learning solution that learns construction scheduling domain knowledge from existing records automatically and uses it to validate logical dependencies in input schedules.

How effective is the method and how is learned knowledge reused?

The paper reports an F1 score of 88.3% for schedule logic validation. It also tailors the method to use learned knowledge to schedule unordered activities.

Oct 22, 2026

Uploaded by Lucas Martin