Intelligent Verification System for Concrete Compliance and QA Risk Management
Digital Asset Delivery and Project Management
Challenge Statement Owner
Expand Construction is one of Singapore’s major home-grown building construction groups with integrated civil engineering and construction support service capabilities. The company is principally involved in the construction of residential, industrial and special niche projects as a main contractor. Expand Construction has built an impressive track record of notable and iconic projects with excellent performance and high-quality products and services to its clients.
Background
Many construction processes involve multiple steps, handled by different people, and recorded in disconnected formats, including paper slips, PDFs, spreadsheets, mobile chats, and siloed databases. These fragmented workflows introduce risks of human error, inconsistent documentation, and delayed decision-making.
Concrete quality management is one of the clearest examples of this challenge. Each pour involves a chain of interdependent actions, from delivery, slump testing, and cube casting to compressive strength testing. These are typically managed by different teams and vendors.
Data from these steps often resides in separate systems or formats, making it difficult to verify whether the concrete meets specification, especially when preparing for progress claims, regulatory inspections, or audits. Issues such as missing or delayed test results, mismatches between delivery orders (DOs), cube IDs, and strength reports, or inconsistent naming conventions can compromise traceability. QA teams spend significant time manually reconciling documents, while finance teams risk certifying claims without full assurance of compliance.
While some suppliers now offer digital DOs or online lab reports, much of the coordination and information exchange, including delivery confirmations, photos, and test results, still happens over WhatsApp. This results in fragmented data trails that are difficult to verify or reconcile. There is currently no intelligent system that automatically processes this multi-source data to ensure compliance or flag risks that could affect progress claims or audit readiness.
The Challenge
How might we build an intelligent system that automatically verifies concrete quality and compliance from delivery through testing, and flags risks that could affect audit readiness, milestone claims, or regulatory approvals?
Requirements
Functional Requirements
- Extract key data fields (e.g., DO number, slump result, cube ID, mix grade, strength value) from a variety of sources, including WhatsApp messages, PDFs, scanned slips, spreadsheets, and eDOs using AI-based document recognition
- Reconcile delivery records, slump test results, and cube strength data by batch or pour, even with inconsistent formats or naming conventions
- Automate workflows that link delivery, testing, and approval records across projects
- Detect and flag anomalies such as missing or overdue test results, strength non-conformance, unmatched or duplicate cube IDs, and incomplete test coverage for pours linked to progress claims
- Present compliance status, traceability gaps, and verification insights through interactive dashboards for use by project, finance, and audit teams
Technical Requirements
Core Requirements:
- AI/NLP engine for extracting data from semi-structured documents
- Reconciliation logic engine to link DOs, slump tests, and lab reports into a coherent QA trail
- Configurable alert system to detect missing data, failed tests, or data mismatches
- Role-based dashboards for QA, finance, and operations users
- Drag-and-drop upload functionality or auto-ingestion from email and chat platforms
Additional Preferences:
- Integration with supplier portals for DO validation
- Integration with finance systems for claim tracking
- Compatibility with Common Data Environments (e.g., Autodesk Construction Cloud)
Expected Outcomes
- Reduction in time and errors in compiling QA documentation for concrete delivery and testing
- Improvement in traceability and confidence during project audits and regulatory inspections
- Reduced risk of unverified concrete strength data being used for progress certification
- Increased confidence in milestone-based payment approvals
- Structured QA data sets that support ESG reporting, analytics, and carbon tracking in concrete usage
- Scalability across infrastructure, commercial, and residential projects with regular concrete activity
- Improvement in productivity, saving up to 2 working days per month otherwise spent collating raw data and verifying them against site works
- Enhanced workflows by freeing QA/QC engineers from mundane, repetitive tasks, enabling them to focus on higher-value work
Deployment Environment and Constraints
The system must be able to handle document and data formats commonly used in Singapore construction projects, including scanned records, supplier-issued eDOs, and lab-issued strength reports. Integration or compatibility with project file storage systems, mobile chat platforms (especially WhatsApp), and shared QA folders is highly desirable. User roles will include site teams, QA/QC officers, finance staff, and auditors.
Proof-of-concept (POC)/Pilot Support
Expand is prepared to run a pilot on an active project involving daily concrete pours. The company will provide:
- Real-world datasets from ongoing projects, including delivery slips, test results, and documentation.
- Access to QA, operations, and finance teams for validation and feedback.
- Assistance from a technical team currently supporting their AI ingestion system, with integration and deployment
- Full site access for pilot implementation and iterative testing
Expand is open to working with early-stage or mature solution providers and can co-develop or integrate the solution as needed.
Commercialisation and Scaling
Upon successful validation, Expand will scale the solution to multiple live projects and integrate it into its broader QA, finance, and audit workflows. The company sees this as a step toward intelligent quality assurance in large-scale construction.
This challenge reflects a broader gap in the industry. Most contractors do not currently have integrated systems for validating ready-mix concrete (RMC) delivery data. A successful solution could be scaled across the sector to standardise QA compliance processes.
