Digital work instructions and Odoo integration at e-power
How e-power connected digital work instructions in Azumuta to Odoo, focusing on assembly, quality records and shop floor support.
Digital work instructions only work if they fit the job on the assembly line. Operators need to know which step comes next, which check is required and where to record the result. At generator manufacturer e-power, this was the focus of a project involving Azumuta and Odoo. We guided the implementation and the integration between the shop floor and ERP.
From an empty hall to digital assembly
e-power started a new production line in an empty hall. Within a few months, it needed to become a digital assembly environment built around quality. New staff had to be trained, while existing staff had to learn to work on a new product. The instructions therefore had to be more than available. They had to be clear and practical to follow.
The need went beyond replacing paper with a screen. Staff needed clear work steps, with quality checks that could be tracked and recorded digitally. Integration with the existing ERP system, Odoo, was essential: work instructions and quality records could not become a standalone application.
Azumuta on the shop floor, connected to Odoo
e-power chose Azumuta for digital work instructions and quality checks. A Vlegel project manager was on site full-time to guide the rollout and integration. That daily presence meant staff questions could be answered, problems addressed and the application kept aligned with the production process.
Behind that project manager was the wider Vlegel team, with expertise in PLC programming, software development and ERP integrations. That matters in projects like this: a question on the shop floor might concern an unclear instruction, but it could just as easily involve data exchange or integration with a technical process.
The Odoo integration enabled real-time data exchange between the systems. The reported outcome was an environment with clear digital instructions and a streamlined quality process. The original case study also mentions fewer errors, better quality control and productivity gains, but provides no measurements to quantify those improvements.
What to define upfront on the technical side
For a similar project, we start with data and responsibilities. Which system manages which information? How is the right instruction selected for an order? Where is a quality result stored? These are integration design questions, not details to discuss after go-live.
Error handling is part of that design too. If a connection drops, it must be clear which records have not yet been sent. Resending them must not create duplicates. Instruction management also needs clear rules for versions, approval and changes. These are general considerations; the source does not describe how they were implemented technically at e-power.
Connecting machines and AI without blurring responsibilities
When machine data is needed too, OPC UA and MQTT can help connect machines. A Unified Namespace (UNS) can make that data available in a shared structure, with recognisable context such as machine, line or order. This is a possible next step, not something demonstrated in this case study.
The same applies to AI. An assistant can read approved instructions and recorded data to find information or summarise deviations. We keep the boundary clear: AI only reads, and a human decides. AI does not change work instructions, sign off on quality or control a machine.
Want to discuss how digital work instructions, ERP and machine data fit together in your production environment? Book a 15-minute conversation with Vlegel Technology in Genk, the software engineering team for machines, data and AI for factories and machine builders. We will look at your current setup and identify which integration or working agreement needs attention first.
