
Boustead Singapore’s project teams were producing reports, but inconsistent information made it difficult to see how projects were performing and spot risks early.
Muhammad Khalil Shaiful Bahari, Boustead Singapore’s Senior Vice President, says the firm is working to make project information more useful for decisions. AI has a role, but only where it addresses a specific problem, and people remain accountable for the decisions that follow.
In this interview with Frontier Enterprise, Khalil explains how Boustead is changing its project controls and where it sees limits to AI’s role.
What project-delivery problems was Boustead trying to solve, and how were these activities managed previously?
One of the biggest challenges we faced was fragmented information. Project data was spread across multiple systems, spreadsheets, emails, and separate reporting processes. That made it difficult to establish one reliable source of project information.
Before we began digitising our processes, much of our project reporting was still manual. Operations and site teams collected updates, prepared reports, consolidated information by email, and reconciled progress in spreadsheets. This took time and meant management often made decisions based on periodic reports rather than real-time information. We were reacting to problems rather than identifying them early.
Consistency was another challenge. Projects could use different reporting formats and methods for tracking performance, making it difficult to see how multiple projects were performing as a whole. Identifying risks also depended largely on our project teams’ experience and periodic reviews. That experience remains important, but we wanted data and technology to help those teams spot emerging issues earlier.
For me, digitisation was never simply about replacing spreadsheets or adding another software platform. We wanted to improve how information moves across the organisation so it could be captured once, shared with the relevant stakeholders, and used reliably for project controls and decisions.
The aim is to move from manual reporting to real-time information about projects, and from reacting to problems to anticipating risks and changes in project performance.
Which AI project-control technologies and platforms is Boustead using on construction sites?
We use project controls that combine building information modelling (BIM), digital twins, cloud-based project management platforms, and data analysis to monitor progress, report on projects, forecast outcomes, and support decisions. We use BIM and site-capture data to help verify progress, automate reporting, forecast costs and manpower needs, and analyse schedules with AI.
This work uses Autodesk Forma, Novade, Microsoft SharePoint, and Microsoft Copilot. By connecting these platforms, we aim to replace manual reporting with real-time project information and identify risks earlier.
What technical or operational challenges arose during implementation, and how did Boustead address them?
One of our biggest lessons was to make sure we were solving the right problem before deciding to use AI. We still need to ask whether a problem requires AI or could be addressed through better processes, data, or existing technology.
Once we had identified where AI might help, the main challenges were connecting data across Autodesk Construction Cloud, Novade, and SharePoint; maintaining data quality; getting people to use the tools; and meeting cybersecurity, privacy, PDPA, and governance requirements. We addressed these by identifying specific problems, involving stakeholders, building prototypes, testing them with end users, and expanding their use only when we could show measurable results.
We also keep people involved in decisions. AI can help us make decisions faster and with more information, but our people remain accountable for those decisions. Ultimately, we judge success by whether AI solves a real problem and delivers measurable value to the business, rather than by how much AI we deploy.
What improvements has Boustead seen in project productivity, reporting, and risk management?
We have seen improvements of at least 20% to 30%, mainly in productivity, reporting efficiency, visibility, and earlier risk identification.
By digitising project workflows and connecting information through Autodesk Forma, Novade, Airsquire, and Wenti Labs, our teams spend less time consolidating information manually, preparing reports, and searching across systems. Project reports are also more timely and consistent, giving management a clearer view of progress and potential problems.
Using 360° site capture and data analysis, our teams can identify discrepancies and risks earlier, reducing unnecessary site walks, rework, and delays. We measure success by whether the technology saves time, reduces manual work, improves decisions, and helps us manage project costs, risks, and timelines, rather than by how much technology we deploy.
Which project-control tasks can AI handle, and where is human oversight needed?
We find AI useful for repetitive tasks involving large amounts of project data. These include consolidating data, generating progress reports, analysing schedules, comparing site progress with BIM models, spotting trends and unusual results, flagging possible delays or risks, and helping forecast outcomes. We have also explored AI, machine learning, and 3D LiDAR scanning to automate construction verification and help teams make decisions faster and with better information.
But AI doesn’t replace professional judgement. People need to review its output whenever it could affect safety, compliance, costs, contractual obligations, or major project decisions. AI might flag something unusual or predict a delay, but the project team must assess the context, check the information, and decide what to do.
AI can handle the analysis, but our people remain accountable, particularly for decisions involving structural safety and compliance.
How will Boustead expand AI across its projects and operations?
I believe the built-environment industry is moving beyond whether to adopt AI and asking where it can deliver value at scale.
In project delivery, we see opportunities to use AI more widely for progress monitoring, automated reporting, identifying risks before they arise, analysing schedules and costs, quality management, and resource planning. We’re also examining how AI could connect information across our common data environment and other project platforms, building on our existing use of AI-assisted 360° site capture and connected project data.
Beyond construction, AI and data analysis could support facilities and building operations, sustainability and energy use, corporate reporting, knowledge management, and other processes across the business.
We also want to scale the agentic workforce by developing the governance and culture needed to increase AI’s impact on the organisation and workforce.
This fits with Boustead’s longer-term move towards connecting building and operational platforms and using data more consistently. We will expand selectively: identify an operational problem, test a proposed solution, show measurable value, put the necessary data and governance controls in place, and then use successful applications across more projects and business functions. Our aim is to use AI where it helps our people make faster, better-informed decisions, while keeping people accountable for those decisions.









