Skip to content

Digitising and automating in-flight Programme Controls Offices & Commercial Offices with OCR + AI

Having presented at the Project Controls Expo on our work with Gardiner & Theobald (G&T) implementing a Power Platform (PowerBi, Power Automate, Power Forms) & SharePoint/O365 reporting solution on TfL’s on the £3bn Piccadilly Upgrade programme, we know the power of new digitising technologies as well as the practical challenges of implementing them in the real-world.

However, through these experiences we have noticed the challenge to implementing this new technology on new projects.

The problem is that despite new technology developments for Project visualisation, data analytics, workflow automation and BI dashboards, many businesses still find it difficult to digitise their project processes as their legacy commercial documents and HSEQ procedures are still in static form. And so, adoption of new project technologies can only happen in piece-meal form, with limited benefit as companies worry of the physical effort in migrating historical data.

However it is precisely these legacy projects where this new technology has the maximum benefits.

The problem with much new digitisation and AI technology is that they can only work on new projects. And so, the major immediate benefit that they can supply to inflight projects, where the actual issues are, is lost. What we need is new technology, to be able to be quickly deployed on legacy projects.”

So how we can help? We can digitise in-flight commercial contracts, perhaps in PDF, through OCR digitisation into a SQL dB to then create Smart AI Templates via SharePoint Forms, ABBYY FlexiCapture, MS Power Apps and Power Automate. We use ABBYY FlexiCapture with Machine Learning Optical Character Recognition (OCR) technology to mass digitise existing Project documentation, commercial current procedures, historical commercial notices, and project records.

Figure 1: High-level solution overview

The result: Migrated historical Commercial and HSEQ records and static PDF procedures can then be migrated into SQL dB to thereon create a BI reporting dB from which we then created dynamic web-based automated procedures integrated with O365.

Figure 2: Key Data considerations

Below we demonstrate how the data elements above are extracted. Once the extracted fields have been migrated to a SQL dB, a 2nd loop would then extract the data into SMART AI templates.

Data once in a SQL dB is then inserted into a Client-Side Pre-Production / Production dB for verification. Click the demo video below to show how we can extract the data for verification, to then import into Power Platform in the 2nd loop to create SMART AI templates.

Figure 3: Demo of OCR extraction of nuclear operating procedures

Q-Screen for intelligent data extraction with high-level of confidence on accuracy.

Database and system checks to confirm against database lists (e.g. Master Equipment Lists, Office 365, DCMS).

The steps here would be to:

  • Design the form fields and connect back to SQL Query Lists.
  • Format templates with correct branding and document references.
  • Ensure form validation based on global parameters selected.

By automating project activities, for example by integrating COINS and Asite with other planning, risk, and cost management systems data to remove manual re-keying of data and refocus projects on benefits realisation rather than form filling. Dashboards and metrics can then be designed to get process blockages alerts on NEC4 compliance, to ensure HSEQ and commercial compliance.

We can then automate the creation of Work Orders to update the Enterprise Asset Management (EAM) system in SAP PM without double entry. Safety inspection can be filled in customised mobile forms remotely, even when there is no mobile signal. Certificates can be submitted electronically to a central data repository and saved in Office 365 and company’s DCMS.

ABBYY FlexiCapture overview

FlexiCapture offers an extensive set of tools to customise the document capture workflow and integrate it into your business process:

  • Customising document capture workflows
  • Customising client station interfaces
  • Support of business rules
  • Using ABBYY FlexiCapture as a web service

The benefits to a Client using ABBYY FlexiCapture are:

  • Client/server architecture with Web Service API
  • Standard workflows for advanced data capture from documents
  • Capable of processing more than 10K pages per month
  • Centralised and distributed capture of documents
  • Automatic classification based on document structure and content
  • Intelligent data extraction from documents of any complexity
  • Conversion to searchable documents for archiving and retrieval
  • Routing of images and data to business workflows
  • Export of images to TIFF, searchable PDF and PDF/A
  • Export of data to files and SQL databases
  • Out of the box connection to Microsoft SharePoint Server
  • Custom export to any business application or ECM system
  • Simultaneous export to of images and data to multiple destinations
Figure 4: Overview of OCR Extraction Forms

Using our Technology Partner, User Friendly Consulting (UFC)’s ABBYY trained and certified personnel, we can create a digital representation of typed textual data from images and documents using a web service API.

How can this be used? To ensure that Engineers are writing their procedures correctly, UFC have classified health care charts (page by page), extracting critical information such as patient related data, and lastly bookmarked specific sections for human review through Workstations as per Figure 5, where ABBYY FlexiCapture is employed for verification of data at the (1) Group Level or (2) Field Level verification.

Figure 5: Human review of OCR extraction through Workstations


  • Classification/Extraction
  • Auditing documents for presence of sections/keywords
  • Machine Learning to accommodate varying locations of data


  • Grammar/Punctuation auditing
  • Translation of documents


  • Verification stations for quality assurance

Document level verification will take the following forms as per the Web Verification Station in Figure 6:

Figure 6: Verification of data through Web Verification stations
Back To Top
Verified by MonsterInsights