Digital Lean Simulation Game Boosts Student Construction Planning and Project Performance

A digital simulation game designed to instruct engineering students in lean construction practices significantly enhanced project workflow reliability, transformed simulated project financial outcomes from net losses into substantial profits, and produced statistically significant increases in conceptual comprehension during validation testing. In empirical trials involving undergraduate civil engineering cohorts, teams transitioning from conventional scheduling to the Last Planner System framework achieved full building completion within designated time limits, elevated their average planning reliability from 60.66 percent to 92.38 percent, and demonstrated marked increases in standardized assessment scores following structured debriefing sessions.

The global construction sector has long grappled with sluggish productivity growth relative to other industrial domains, routinely plagued by persistent project delays, severe budget overruns, and systemic process inefficiencies. To counteract these operational vulnerabilities, project delivery organizations have increasingly turned toward lean construction paradigms, which adapt core tenets of lean manufacturing initially formulated by Taiichi Ohno at Toyota. Rather than relying on traditional top-down commands that risk excessive waste and idle inventories, lean construction concentrates on maximizing delivered customer value while eradicating non-value-adding activities. The methodology rests upon five foundational principles: defining value from the client perspective, identifying the end-to-end value stream, establishing seamless continuous workflow, enabling downstream pull scheduling driven by actual demand, and rigorously pursuing operational perfection by mitigating labor, spatial, and temporal errors.

Among the tools created to operationalize lean methods within construction environments, the Last Planner System, originally introduced by Glenn Ballard and Gregory Howell, functions as a production control system that converts scheduled intentions into executable field tasks. Operating within the Lean Production Delivery System, the Last Planner System organizes production across five interconnected delivery phases: project definition, lean design, lean supply, lean assembly, and final use. The operational core of the Last Planner System relies on a structured sequence that systematically transitions activities through four stages: identifying what should be done via milestone and master schedules, analyzing what can be done through constraint identification and lookahead screening, securing frontline subcontractor commitments on what will be done through weekly work plans, and performing root-cause analysis on what did happen to facilitate continuous learning.

Despite documented advantages across safety, cost mitigation, environmental impact, and schedule adherence, widespread adoption of the Last Planner System across construction enterprises remains hindered by conventional mindsets and insufficient familiarity among site practitioners. Traditional passive pedagogical approaches, predominantly reliant on unidirectional lecture models, have proven ineffective at imparting the dynamic problem-solving and multi-party negotiation skills required for lean implementation. In response, academic researchers have investigated active learning interventions, specifically serious games that simulate authentic industrial environments without exposing learners to actual commercial liabilities. While earlier educational board games such as VILLEGO and LEBSCO successfully introduced physical round-based modeling with toy building blocks, their reliance on physical components and rigid classroom constraints highlighted the need for accessible, fully digital pedagogical tools capable of simulating real-world digital workflows.

To overcome the operational limitations of physical simulation games, the research team formalized the development of LPS-UP using the ReDIRE framework, a five-stage methodology comprising Requirement analysis, Design concept, Implementation, Research framework, and Evaluation. During requirement analysis, researchers confirmed the pedagogical necessity of interactive digital tools for lean training and outlined necessary role structures. The design concept phase established core gameplay mechanisms, role definitions, and digital software integration, subjected to scrutiny through expert validation involving experienced lean construction academics, field managers, and serious gaming scholars. The implementation phase refined system parameters through small-scale pilot testing with undergraduate cohorts, while the research framework and evaluation stages established structured pre-game briefings, comparative experimental rounds, comprehensive knowledge assessments, and moderated debriefing mechanisms.

Architecturally, LPS-UP integrates three commercially available software platforms to replicate the communication, management, and physical assembly dimensions of modern construction management. Players execute structural assembly within SketchUp Web, manipulating a 3D digital model of a Pre-Engineered Building warehouse across designated spatial zones, including active construction areas, material stockyards, and equipment bays. Because the design platform operates on a single-user file-sharing paradigm, simulated site progress is recorded in a Main Game File stored in the skp file format, which is systematically exchanged among trade contractors via the Zoom meeting communication channel. Concurrently, collaborative production control is conducted on a shared Miro digital whiteboard, which features dedicated interactive panels for pull-planning sticky notes, lookahead constraint tracking logs, weekly progress tracking charts, and structured variance analysis registers.

The complete educational sequence encompasses a five-hour pedagogical curriculum structured into five discrete modules: an introductory thirty-minute lecture and pre-game knowledge assessment, a forty-five-minute mock familiarization round, a seventy-five-minute non-LPS control round, a comprehensive one-hundred-and-thirty-minute Last Planner System round, and a concluding twenty-minute post-game assessment paired with a reflective debriefing. Within each gameplay round, teams simulate eight consecutive work weeks, with each single operational day calibrated to precisely one minute and each complete work week spanning six minutes of active assembly. Project teams operate with eight assigned functional roles, including an operational manager overseeing site sequencing, a procurement manager handling material requisitions, and six trade subcontractors managing specialized packages encompassing demolition, substructure, concrete slab work, steel erection, paneling, and architectural detailing.

The experimental trial evaluated two separate five-member student teams drawn from the civil engineering undergraduate cohort within the Faculty of Engineering at Universitas Indonesia. In these experimental sessions, team configurations were adapted to five participants by consolidating contractor roles, supported by three dedicated facilitators acting as process moderators, material suppliers, and independent metric accountants, alongside an overarching master facilitator. In both rounds, teams faced identical structural engineering challenges, complete with randomized weekly uncertainty cards introducing realistic site contingencies such as severe weather, equipment breakdowns, or material discount opportunities, thereby isolating the planning methodology as the primary independent operational variable determining project execution performance.

Under the conventional non-LPS operational paradigm, wherein participants coordinated solely through a static master schedule with bi-weekly updates and restricted real-time communication, both student cohorts failed to complete the warehouse structure within the forty-eight-minute aggregate operational limit. Team A completed only 63.95 percent of the building assembly after severe execution delays in concrete slab casting during week three cascaded into subsequent structural steel delays, ultimately stranding the project with zero progress across paneling and detailing packages during weeks seven and eight. Team B achieved 70.07 percent completion, similarly bogged down by sequential bottlenecks in foundation and framing operations that left wall paneling only partially addressed and final detailing untouched. Financially, both groups accrued heavy penalties, concluding the non-LPS round with substantial net financial deficits of negative 13,170 currency units for Team A and negative 12,093 units for Team B.

Conversely, the introduction of the collaborative Last Planner System framework during the second operational round catalyzed an immediate recovery in production flow and project financial performance. Armed with interactive pull-planning whiteboards, lookahead constraint analysis, and mandatory weekly coordination meetings, both student teams achieved 100 percent structural completion well before the final time cutoff, registering total construction durations of 44:46 for Team A and 44:30 for Team B. By mitigating equipment rental overruns and avoiding contractual default penalties, Team A reversed its operational deficit to secure a net project profit of 19,470 units, while Team B generated a net profit of 27,800 units with zero incurred penalty costs. Furthermore, enhanced visibility enabled teams to exploit workable backlogs, smoothly shifting unconstrained future tasks into current cycles when labor capacity permitted.

Quantitative analysis of weekly Percent Plan Complete metrics confirmed a profound stabilization of workflow reliability between the experimental phases. While average planning reliability hovered at 55.79 percent for Team A and 65.53 percent for Team B under conventional planning—yielding a combined baseline average of 60.66 percent—LPS implementation drove average reliability to 94.44 percent for Team A and 90.32 percent for Team B, establishing a combined operational average of 92.38 percent. Standardized knowledge evaluations administered before and after the game demonstrated corresponding intellectual gains; across all ten participants, the mean score on a ten-question conceptual quiz increased from 4.4 to 7.7. Statistical evaluation via a two-sided Wilcoxon signed-rank test for paired samples confirmed that this knowledge improvement was statistically significant (z = 2.72, p = 0.0066, effect size r = 0.91), while the median score shifted from 5.0 to 8.0.

During post-game debriefing dialogues, participants articulated that transparent weekly commitments and explicit constraint removal mechanisms were pivotal in maintaining momentum, contrasting the collaborative environment against the frustration and confusion experienced during uncoordinated non-LPS scheduling. However, the study authors explicitly noted several critical methodological limitations that restrict broad external generalization. Because the pedagogical design utilized a fixed operational sequence where the conventional round always preceded the lean round, observed performance improvements may have been partially influenced by natural player learning curves and increasing software familiarity. The authors also underscored the narrow sample size of ten participants, the intensive facilitator staffing requirement, and the friction caused by operating three unintegrated software platforms, recommending that future investigations implement dedicated control groups playing repeated non-LPS rounds alongside unified custom simulation software to definitively isolate pedagogical mechanisms.

Subject of Research: Evaluation of the digital simulation game LPS-UP for teaching lean construction and the Last Planner System

Article Title: LPS-UP: A digital simulation game for teaching the Last Planner System as a lean construction method

Article References: Machfudiyanto, R. A., Rachmawati, T. S. N., Jauhari, D. U. P., Freese, M., Rahmawan, A., & Verbraeck, A. (2026). LPS-UP: A digital simulation game for teaching the Last Planner System as a lean construction method. Heliyon, 12(16), Article e45675. https://doi.org/10.1016/j.heliyon.2026.e45675

Image Credits: AI Generated

DOI: 10.1016/j.heliyon.2026.e45675

Keywords: lean construction, Last Planner System, simulation gaming, active learning, construction management, engineering education, percent plan complete, collaborative planning, lookahead planning, pull planning, workflow reliability, SketchUp Web, Miro Board, production control, project scheduling, ReDIRE framework, educational games