Manufacturing · AI · Digital transformation
Is Your Factory Ready for Claude AI Agents?
Before asking what AI agents can do for your factory, ask whether your data is ready for them.
There is a lot of discussion about AI Agents—including Anthropic’s Claude for Manufacturing announcement—and how they are going to change the way manufacturing businesses operate.
We have been working with manufacturing companies for nearly a decade on their digital transformation. Unlike many other industries, a significant portion of manufacturing data is still on-premise. It is spread across multiple applications, databases, servers and different formats.
Some factories have ERP, MES, APS, QMS and other applications. Some still depend on Excel sheets and paper records for their day-to-day operations.
Even when data is available digitally, it may not be accessible through APIs.
So, before we talk about bringing AI Agents into manufacturing, I think we need to answer a few basic questions.
10 questions to assess your factory's AI readiness
Let's look at ten important manufacturing capabilities.
| Manufacturing capability | Key question |
|---|---|
| Production Planning | How is your production schedule created and maintained? |
| Production Monitoring | How do you capture actual production against the plan? |
| Downtime Analysis | How are machine downtime events captured and classified? |
| Predictive Maintenance | Do you have access to machine health data and maintenance history? |
| Quality Analysis | How are quality inspections, defects and rejections recorded? |
| Energy Optimisation | Can you track energy consumption at machine or production-line level? |
| Inventory Intelligence | Can you get real-time visibility of raw material, WIP and finished goods? |
| Delivery Prediction | Can you track order progress and predict completion and delivery dates? |
| Root Cause Analysis | Can you correlate machine, production and quality data to identify root causes? |
| Operational Reporting | Can you get a consolidated view of factory performance without manually compiling reports? |
For every question, identify three things:
- Which application or data source holds the information? ERP, MES, APS, QMS, CMMS, EMS, SCADA, Excel or Paper?
- Is the data collected manually or automatically?
- How accessible is the data to other applications?
How do you score your factory?
For each of the ten capabilities, assess how the underlying data is maintained and accessed.
| Data availability and accessibility | Score |
|---|---|
| Manual records / Paper-based | 0 |
| Digital records, but no API access | 5 |
| Digital applications with custom / non-standard API access | 8 |
| Standard applications with accessible APIs | 10 |
The maximum score is 100.
What does your score tell you?
| Overall Score | Assessment |
|---|---|
| 80–100 | Strong digital data foundation |
| 60–79 | Good digital foundation, but integration gaps exist |
| 40–59 | Partially digitised |
| 20–39 | Significant digital gaps |
| 0–19 | Fundamental digitisation required |
What should you work on next?
80–100: Strong digital data foundation
You already have a good digital foundation. Most of your manufacturing information is available through applications and APIs.
The next step is to connect these applications, establish common data models and make the information meaningful across different manufacturing functions.
60–79: Good digital foundation, but integration gaps exist
You have implemented digital applications, but data is still distributed across different systems.
Focus on application integration, API standardisation and eliminating data silos.
40–59: Partially digitised
Some manufacturing functions are digitised, while others still depend on spreadsheets, manual records or disconnected applications.
Focus on digitising critical processes, automating data collection and making existing digital records accessible.
20–39: Significant digital gaps
A considerable amount of manufacturing information is still maintained manually or in isolated digital records.
Start with structured data collection, machine connectivity and digital applications for critical manufacturing functions.
0–19: Fundamental digitisation required
The first priority is to move away from paper-based processes and establish a digital foundation.
Start digitising production, quality and maintenance records. Identify the critical machines that need to be connected and build your digital roadmap.
Is data accessibility enough?
Not really.
Even if all your manufacturing applications have APIs, an AI Agent still needs to understand the context.
For example, a machine breakdown is not just a maintenance event. It can affect production schedules, WIP, delivery commitments and operational costs.
An AI Agent needs to correlate information across these applications to understand the actual impact.
I see three important requirements:
1. Connected Data
Machine-level information and application data should be accessible, irrespective of whether they are hosted on-premise, in a private cloud or on public cloud infrastructure.
2. Contextual Data
Information from ERP, MES, APS, QMS, maintenance and energy applications should be understood in relation to each other.
3. Actionable Intelligence
AI Agents should be able to analyse information, identify problems, recommend actions and, where authorised, execute actions through connected applications.
This also requires reliable data, security, permissions and proper integration with existing systems.
Where do we go from here?
I believe AI Agents will create significant opportunities in manufacturing.
But I don't think the first step is buying another AI platform or replacing the applications that factories have already invested in.
We need to look at what data is available, where it resides, how it is collected and how accessible it is.
AI Agents are evolving quickly. Manufacturing needs to catch up, not by buying more software, but by making the data and applications it already has work together.
Before asking what Claude AI Agents can do for your factory, ask yourself whether your factory's data is ready for them.
Assess Your Factory's AI Readiness
We have put together a practical Manufacturing AI Readiness Assessment covering these ten manufacturing capabilities.
Download the assessment, answer the ten questions and calculate your factory's score out of 100.
Use it to identify your digital gaps and understand what needs to be addressed before introducing AI Agents into your manufacturing operations.