Babcock International

The numbers tell the story clearly. Total validation time for 404 IFC files dropped from 128 hours with the traditional method, to 60 hours with the semi-automated Solibri approach, to just 13 hours with Morta and IFC Open Shell. A process that used to consume 16 working days now finishes in half a day.

Per model, validation went from 19 minutes to 2 minutes. Per cycle, digital engineers save 114 hours of manual effort — time they can now spend on higher-value work like design coordination and quality assurance. Engineers no longer have their workstations locked up running Solibri checks; the entire process runs unattended on virtual infrastructure.

But the real impact goes beyond time savings. Babcock is the appointing party — unlike most organisations using Morta, they specify requirements and maintain assets. Automated, continuous validation means they have consistent confidence in the data flowing into asset management systems. The team is now exploring risk, opportunity, and cost analysis through automation, validation of information delivery plans using Python, and enhanced security checks for information requirements. The foundation they have built keeps opening new doors.

Devonport in Plymouth is the largest naval base in Western Europe and the only place in the UK that can maintain the Royal Navy’s submarine fleet. Babcock is driving a massive infrastructure overhaul there — £2–3 billion of capital expenditure over the next ten years, with approximately 1,300 members of staff on an enduring programme.

At the heart of the challenge is data validation. Every IFC model delivered by the supply chain needs its alphanumerical data checked against information requirements. The numbers are staggering: 404 IFC files, each with 1.2 million lines of data, totalling 4.8 billion lines that need validating.

The traditional process meant downloading IFC files from the CDE, opening Solibri, loading the files and rule sets, running checks, and exporting results. Nineteen minutes per model — and the user’s machine was completely locked up the entire time, unusable for anything else. Across the full set of 404 files, a single validation cycle took 128 hours. That’s 16 working days. On a billion-pound programme with frequent model updates, that simply was not sustainable.

The breakthrough was realising they could eliminate Solibri entirely. Instead of running a desktop application that locks up a machine for 19 minutes per model, they could specify information requirements in Morta, fetch them via the API into Python scripts using IFC Open Shell, and validate data programmatically — reducing validation to two minutes per model.

Babcock got there in three stages. The first was the traditional Solibri approach: download, open, load, run, export. Nineteen minutes per model, user locked out. The second stage introduced virtual desktop infrastructure so validations could run in the background — Python scripts triggered Solibri’s auto-run, results went to Power BI, and users were free to keep working. That brought it down to nine minutes per model.

The third stage eliminated Solibri altogether. Information requirements live in Morta as structured data. The Python script pulls those requirements via the API, validates each IFC file against them using IFC Open Shell, and outputs pass/fail results in seconds. The system even fetches project-specific requirement tables automatically based on IFC metadata, so each project gets the right validation criteria. Requirements evolve as projects mature through design and construction phases, keeping validation relevant at every stage.

The whole pipeline runs unattended: IFC files land in the CDE, sync to a central server, get validated against requirements, and results are pushed into Power BI dashboards and emailed to stakeholders automatically. Continuous, 24/7 validation with no human intervention.

What makes Babcock’s story unusual is their position in the supply chain. Most Morta users are appointed parties delivering information. Babcock is the appointing party — they specify the requirements and then validate what comes back against those requirements. That dual role means they know exactly what data they need and can validate it rigorously against their own standards.

The technical implementation centres on a Python script workflow: pull information requirements from Morta’s API, check data in IFC files against those requirements using IFC Open Shell, output pass/fail results. Project-specific requirements are fetched automatically based on metadata in the IFC file itself, so there is no manual configuration per project. The automated pipeline — CDE to central server to validation to Power BI to email — runs continuously without human intervention.

The team did not jump straight to the final solution. They evolved through three generations, learning at each stage what could be automated further. That iterative approach — proving value at each step before pushing further — was key to building confidence and buy-in across the programme.