Introduction

Within supply chains data can be held in a variety of formats, at different levels of detail, and employing different semantic models. This often makes it challenging to combine data from different sources into a coherent and cohesive form to support the innovation in goods and services.This interoperability is a key cross-sector challenge in situations where data must be exchanged between different digital systems. 

In addition, these systems are often owned and operated by different businesses with often competing interests, which makes it very difficult to determine how to fairly share the risks, benefits, and obligations of any interoperability solution.

This interoperability flagship project focuses on the rail construction industry as they look to deliver the Integrated Rail Plan for the North and Midlands. The ambition is to create a ‘Google Translate’ for the rail construction supply chain, allowing information to be translated semi-autonomously so that disparate software systems and teams can effectively communicate and work together. 

The project work will be carried out with continual monitoring and feedback from key industry stakeholders, which are Network Rail, Bentley Systems, Mott MacDonald, Jacobs and Costain. The project work itself will be carried out by NPL, WizDev with key support from Bentley Systems, and will be led by Digital Catapult.

Why is this important?

One of the key blockers for effective information flow across complex systems is the problem of interoperability of the systems between different nodes in the supply chain. The process of data exchange between different nodes within the supply chain is often slow, expensive, and may even exclude key information to the extent that it makes the exchange superfluous because the data is stored in myriad different ways.

AI and machine learning possess the power to streamline data exchange, enabling a seamless flow of information, thereby unlocking the inherent value within the data and improving decision-making. Through the development of suitable business and collaboration models, active engagement of stakeholders becomes a driving force in ensuring the successful large-scale deployment of this transformative technology. 

What will the project involve?

The project will be delivered through a technical stream and a business stream, both of which are crucial for making the project a success.

The technical stream will develop proof-of-concept AI/ML tools that will convert data from  source data models to target data models. The tools will support conversion between many possible source and target data models by using a common meta-ontology, to be developed during the project, that will serve as the common language for interoperability. 

A key goal will be make to the AI/ML and meta-ontology highly adaptable and extensible so that the interoperability solution can be grown on an agile, ‘as-needed’ basis, potentially as part of a common, open platform and as per the requirements of stakeholders who adopt the solution at different points in time. 

The proof of concept we aim to develop will crucially enable all participating systems to be both data generators and consumers, laying the groundwork for a connected data environment for all parties – subject to appropriate permissions.

Central to the success of this project will be the development of business and collaboration models enabling value generation which we will explore during the project. The emerging models will form the business blueprint for scaling up the POC into a shared utility within digital supply chain applications, where all stakeholders and data owners can share rights and obligations in a fair and balanced way.