Digitizing the On-Site Decision Process: From Experience to System

How business interviews and mental model mapping guided Kashima Construction’s upcoming corporate DX initiative.

Project Overview

• Company: Kashima Construction
• Project: Construction Implementation Planning System (PoC)
• Role: Product Designer (UX/UI)
• Period: Jun 2025 — Aug 2025
• Team: 1 Product Designer, 1 Algorithm Engineer, 2 Project Managers
• Scope: User Research (Interviews), Workflow Design, System Concept, Wireframes, Prototyping

1. Introduction

Construction planning at Kashima Construction heavily relied on the tacit knowledge and experience of senior site managers. Scheduling, resource allocation, and real-time adjustments were managed through uncodified information, making it difficult to transfer critical operational skills to junior staff.

As the sole product designer for this PoC, my role was to design their first construction planning system. Instead of simply digitizing paperwork, the objective was to understand the unwritten operational logic from the managers’ heads and translate it into a structured digital interface.

2. Challenge: The Reality of On-Site Operations

The initial corporate mandate was to standardize the construction planning workflow to accommodate increasing work-hour constraints and overtime limits. However, conventional enterprise UI patterns do not function in this domain.

Through initial research, I found that site operations are highly fluid — daily worker counts fluctuate constantly. What site managers actually prioritize is not a rigid mathematical schedule, but the following practical variables:

• Can I secure the required number of workers for critical construction phases?
• What does the resource allocation curve look like across the project timeline?
• Where are the peak labor bottlenecks, and how can we smooth out those peaks?

Standardization could not be achieved by imposing an abstract system ideal. It required mapping the interface directly to the specific decision-making hierarchy used on-site.

3. Process: Researching the Timeline of Information

To define the system’s role and scope, I focused on how the required density and scope of information transform over time. The operational context and the data needed shift drastically depending on the project phase.

To validate this, I conducted interviews with project stakeholders and analyzed existing planning documents to clarify role-specific responsibilities. I then created low-fidelity mockups using Figma Make to serve as a collaborative baseline for discussion. By iterating through these wireframes with site managers, we successfully defined the exact UI requirements and data structures for different milestones — specifically differentiating between the strategic view needed two months prior to construction and the tactical view needed for day-to-day operations.

4. Design Strategy: Minimizing Complexity via OOUI

To ensure durability in high-pressure field environments, I kept the system features minimal and versatile. I applied Object-Oriented User Interface (OOUI) principles, anchoring the layout around core operational objects (such as workers and site assets). This structure keeps the information architecture stable whether users are conducting long-term planning or making quick daily updates.

Our team mapped the system into four task-based scenarios to align stakeholders. A key focus was balancing algorithm optimization with human intervention:

• Human-in-the-Loop Workflow: After the algorithm calculates an optimized schedule based on deadlines, the system prompts a manual review. The UI visually surfaces the specific timeframes where worker and material constraints must be secured, allowing site managers to manually adjust allocation curves based on real-time variables.

5. Impact and Validation

During the final presentation to decision-makers and site managers, the scenario-based design received strong validation, with users stating, “This is something we can actually use.”

The turning point was a feature driven directly by our business interviews: site managers frequently need to communicate specific instructions to different subcontractors. We designed the UI so that managers can filter information by subcontracting companies or individual craft types with a single click. The system instantly translates these filtered lists into a visual site diagram that displays the exact sequence of tasks.

By aligning the UX with real-world mental models, the PoC achieved the following outcomes:

• Stakeholder Alignment: Established a shared, definitive understanding of construction planning workflows across corporate and site stakeholders.
• Operational Efficiency: Identified clear opportunities to reduce management staffing needs at small-scale construction sites, earning endorsement from site managers.
• Project Advancement: Sparked positive, factual discussions toward the next development phase, with a full-scale corporate rollout currently planned from 2026.

Key Takeaways

1. Understand the Business Flow Before the Screen: Building a system is never the goal. Product design must map the current reality (As-Is) to define a pragmatic future workflow (To-Be) that genuinely improves the business.
2. UX Must Fit the Context: A theoretically perfect system is useless if it does not align with the user’s situational mental model and operational timing.
3. Embrace Subtraction: In complex B2B environments, clearly defining system boundaries and maintaining minimalist, object-oriented design prevents over-engineering and ensures technical feasibility.