Why AI Is Changing the Economics of Manufacturing Systems Integration

Introduction

For many manufacturers, opportunities to integrate systems and automate processes have existed for years. I've been working on them most of my career.

The problem has rarely been identifying what should be improved. More often it has been justifying the cost of doing it.

What we are finding is that AI is changing that calculation.

Not because software development has become obsolete, but because the effort required to build and maintain integrations, workflows and bespoke applications is falling.

We are expecting an uptick in these smaller projects as these quicker wins.

---

Manufacturing Transformation Is Often Incremental

The popular image of digital transformation is often one of large programmes, major system replacements and wholesale organisational change.

Our experience is somewhat different.

Most manufacturers and others we work with are engaged in a continuous programme of smaller improvements. They are looking to improve operational visibility, reduce manual effort, control costs and provide better information for decision-making. Some projects are strategic, others tactical, but collectively they contribute to the ongoing development of the business.

Many manufacturing businesses have invested in specialist software over the years. Some systems have been developed in-house, others supplied by specialist vendors, and still others delivered by software partners. These systems often support critical processes and are deeply embedded within day-to-day operations.

As businesses evolve, additional requirements emerge.

Increasing levels of product customisation create new data requirements. We've been pulling in for example Drop Ship requirements from end clients into manufacturing plans to provide early labelling. Operational improvements require better visibility of time and activity. Commercial pressures demand more accurate costing and pricing information.

Meeting these needs often requires lightweight applications, bespoke workflows and integrations between existing systems.

Historically, the challenge has not been identifying opportunities. It has been building the business case to deliver them.

Integration Has Always Been Important

Most manufacturers already possess a substantial technology estate.

Typical organisations may operate:

• ERP systems
• Manufacturing systems (sometimes called MES or even MIS)
• Warehouse systems (WMS)
• Quality management platforms
• Finance software
• Reporting tools
• Departmental databases and spreadsheets

Individually, these systems often perform their roles effectively.

The challenge is and has for a while been what happens between them.

Information is frequently entered multiple times, reports are created manually and employees spend valuable time moving information between systems that were never designed to work together. Sometimes the human making a small semantic shift between the product name in production to the product numbers in distribution for example.

These gaps are often accepted because the effort required to remove them can be significant.
Many organisations have identified opportunities to automate these processes for years. The limiting factor has often been the cost of implementing the required software.

---

Where AI Is Making a Difference

What we are seeing is that AI is influencing this transformation in two important ways.

---

Internal Teams Can Deliver More.

The first impact is within manufacturing businesses themselves.

This is not an entirely new phenomenon.

For decades, SMEs have relied upon capable and enthusiastic individuals who developed practical software solutions using tools such as Microsoft Access, FoxPro or dBase. Many businesses benefited from these systems because they enabled customisation and automation at a cost that would otherwise have been difficult to justify.

Often these solutions were created by power users who understood the business problem and had enough technical capability and motivation to build something useful.

AI-assisted development is extending this capability.

We see organisations use technically minded employees to prototype applications, automate workflows and develop lightweight systems using modern technologies such as Python. A couple of weeks ago I saw a great looking little product free standing customisation system written by an in-house enthusiast at a client, we have added a link to pull some of this data into a labelling system.

This does not mean everybody has suddenly become a software engineer.

It does mean that individuals with domain knowledge and some technical aptitude can often achieve more than would previously have been possible.

As a result, manufacturers can address smaller operational challenges that may previously have remained unresolved, moving the company slowly forward in weeks and months rather than years.

 

---

Professional Development Has Become More Efficient

The second impact affects organisations like ours.

Despite some of the headlines surrounding AI, business-critical systems still require significant expertise.

Software must still be specified, designed, tested, secured and maintained. Integrations still need to be reliable, supportable and aligned with operational requirements.

What is changing is the productivity of experienced development teams.

Historically, a substantial proportion of project effort was spent writing the core code.

Today we are finding that the balance is shifting.

The actual production of code is becoming a smaller part of the overall process. More emphasis is being placed on requirements gathering, solution design, testing, governance and validation.
In simple terms, our developers are spending less time typing code and more time applying professional judgement.

This is an important distinction.

A degree of what has become known as "vibe coding" may be perfectly acceptable for organising personal tasks or producing a quick prototype. For systems that influence manufacturing operations, costing, logistics or customer orders, organisations still need confidence that the software is robust and supportable.
AI works best when combined with experienced professionals who understand software architecture, integration, security and long-term maintainability.

---

The Economics Are Changing

One manufacturer we work with illustrates the trend well.

The business produces products to exacting physical specifications and has seen increasing demand for product customisation over recent years. Whilst the physical products themselves have changed relatively little, the information surrounding them has become significantly more complex.

To support this, the company has been developing a range of smaller digital initiatives. Some focus on improving manufacturing control and costing. Others help manage the growing volume of customisation data flowing through ordering, production, delivery and labelling processes.

None of these projects would normally justify the description "digital transformation" on their own. They are simply practical improvements that solve specific operational problems.

What is interesting is that several initiatives which might previously have remained on a wish list are now moving into delivery. Internal staff are building prototypes and lightweight applications, whilst specialist software providers like us can deliver integrations and bespoke systems more efficiently than was possible a few years ago.

Collectively, these projects are helping the business make steady progress through a series of smaller, commercially justifiable improvements.

---

More Opportunities for Incremental Improvement

As the economics improve, we expect to see more investment in smaller improvements.
Examples might include:

• Connecting existing systems more effectively
• Automating repetitive administration
• Improving manufacturing data collection
• Enhancing costing and pricing visibility
• Reducing manual rekeying of information
• Supporting product customisation processes
• Building lightweight operational applications
• Improving reporting and decision support

Individually, these projects may not transform a business.

Collectively, however, they can have a significant impact on efficiency, visibility and profitability.
This accelerates what we have always seen on the ground.

For most businesses, success doesn't so much come from a single knock out transformative project. More often it comes from a continual process of refinement and improvement.

---

What Should Manufacturers Be Asking?

Much of the discussion around AI focuses on entirely new capabilities. We are looking at these for clients and might discuss them elsewhere but an equally valuable question may be simpler.

• Are there projects that were previously rejected because the costs outweighed the benefits?
• Are there manual processes that everyone accepts because addressing them never quite made financial sense?
• Are there systems that should be integrated but have remained disconnected because the investment required was difficult to justify?
• Are there opportunities for technically minded employees to solve more problems internally?
• Are there specialist integration projects that merit a second look?

If the cost of delivering software is falling, some of those answers may now be different.

---

Conclusion

For manufacturers, one of the most immediate may be the changing economics of software development itself. This is probably true in other sectors, but in our client base this is the first place we are noticing it.
The need for expertise has not disappeared. Effective software still requires careful design, rigorous testing and experienced delivery.

What has changed is the cost-benefit balance.

As software becomes more affordable to create and integrate, manufacturers gain the opportunity to make more incremental improvements, solve more operational problems and extract greater value from their existing systems.

The question is no longer simply whether an improvement is possible.

For many manufacturers, the next phase of digital transformation may not come from replacing everything. It may come from finally tackling the smaller improvements that never quite justified the investment before.