GUIDE · MANUFACTURING OT

Updating PLCs and SCADA: maintenance windows and trade-offs

By the end of this guide you will know how to decide whether and when to update the controllers and supervisory systems in a plant. You will see the prerequisites to check, the steps for planning an intervention without surprises, the typical mistakes, and the measures to use when an update is not possible.

Updated october 2026
6 min read
Altovar Technical Editorial Team

1. Why updating is harder in a factory

In an office, an update is applied overnight and nobody notices on Monday. In a plant, an update can change how a machine behaves, invalidate a manufacturer's approval or require the line to be stopped for the time needed to bring it back into operation. The controller may also be programmed with a development tool that must be in a compatible version, which adds a dependency that does not exist in the office.

For this reason the question is not whether to update, but how to do it without compromising the process. The first step is to understand which of the two matters more for the line. Known vulnerabilities in older devices often remain open for years, which is a common situation. The useful question is which risk stays open, for how long, and with which measures that reduce it in the meantime.

2. Prerequisites before planning

A planned update starts from some information that must be available before you ask production for a window. Without it the window becomes a gamble, and the line pays for it. Before talking about dates, gather the following information and assign it to a specific person, who keeps it up to date and hands it to the supplier. Assign it to someone who knows both the machine and the network.

  • List of controllers and supervisory systems, with current version
  • Written confirmation from the manufacturer that the target version is supported
  • Verified copy of programs, configurations and licences
  • Programming tools in the correct version, already available
  • Written and tested rollback procedure

3. The steps for a controlled update

The steps below must be carried out in the order given. The most important point is testing: an update that has not been tried on a copy or on a test bench is a risk you have chosen not to see. Each step must leave a written trace, with date, person and result, because it will be the reference if something does not add up. Keep the traces in the same archive as the plant documentation.

Do not skip the final check. After the update, verify that alarms, communications between controllers and panels, and safety functions behave as before. A system that appears to work may have lost a communication that will only surface at a critical moment, when the line is already at full production. If something does not add up, do not declare the intervention closed, and go back to testing.

  1. Verify with the manufacturer that the version is approved for that machine.
  2. Read the release notes and identify the changes that affect your process.
  3. Carry out a full backup and verify that a restore is possible.
  4. Test the update on a copy of the system or on a spare machine.
  5. Agree an intervention window and a return plan with production.
  6. Apply the update following a written procedure, with a competent person present.
  7. Check the line operates correctly and record the installed version.

4. Typical mistakes

The first mistake is applying an update simply because a new version has been released, without checking that the manufacturer supports it for that device. The second is underestimating the programming software, which may require a different version of the controller and make a quick restore impossible. The third is not having a copy of the program in the correct format, kept with the documentation and not only on the computer that failed.

Another common mistake is planning the update without the supplier who knows the machine, or relying on a single person for the entire window. If something goes wrong, you need someone who can decide to roll back without hesitation. Always arrange a second person on call and a clear criterion that says when to stop the intervention. The same applies to anyone updating without having read the release notes.

5. When you cannot update

Some devices cannot be updated: the manufacturer no longer supports the version, or the update would require a replacement that is not in the budget. In these cases the risk is not closed, but reduced. The most useful compensating measures are isolating the device in a separate zone, limiting who can access it, blocking unnecessary services and controlling the flows that reach it.

Every exception must be recorded with the reason, the person accountable and a review date. An exception without a review becomes permanent without anyone deciding so. When the situation changes, for example with a scheduled replacement, the exception must be formally closed and not simply forgotten. Ask for the review to be entered in the maintenance calendar, so it stays under control even when people change.

6. How to get started

This week, list your controllers and supervisory systems with their current version and the date on which the manufacturer declared support ended. You do not need to know which updates still exist: you need to understand which devices are in a risk situation. Rank them by importance to production, starting with those whose failure would stop the most lines.

Then ask your maintenance supplier, for each of the most critical devices, whether a supported version exists and what steps it requires. With these answers you can propose the first planned intervention to production, with the date, window, test and return plan already written. Also ask how many people are needed and which skills must be present. A complete proposal is worth more than a generic request for a shutdown.

Frequently Asked Questions

Can I update a controller without stopping the line?

In some cases yes, but it is the exception. Many updates require the controller to be stopped, because the program is replaced or restarted. Ask the manufacturer whether the work can be done with the machine running, and plan accordingly. Even when a full shutdown is not needed, it is still advisable to agree a window with production.

How often should updates be applied?

There is no single cadence. It depends on the risks declared by the manufacturer, the availability of windows and how critical the machine is. A good practice is to set a periodic review of the device list and their support status, so that decisions are taken in good time and not in an emergency.

Should the maintenance supplier be involved?

Yes, and from the start. Often only the supplier knows the exact versions and software dependencies. The contract should clarify who does what during the intervention, who authorises the rollback and who is responsible if the machine does not restart. Check the details of liability with a legal adviser.

Do I need a test on an identical machine?

The best test is on a faithful copy of the plant or a spare machine with the same configuration. If none exists, a test bench with the same components is a good alternative. When neither option is possible, the risk of the update increases and must be discussed explicitly with production management.

What is worth recording after an update?

Record the installed version, the date, the person who carried out the intervention, the tests performed and any anomalies. Also keep the copy of the program before and after the change. These records are needed for restoring, for later checks and to demonstrate that the intervention was carried out in a controlled way.

About this content

General informational guide. Regulations and technical requirements change over time: always check the official sources and assess your specific case with a qualified technical or legal advisor.

GET IN TOUCH

Talk to our team.

We'll show how Altovar integrates into your infrastructure. No commitment, no sales pitch. Just a technical conversation.

BOOK A DEMO →