If you want to implement equipment monitoring, it’s best not to buy any software until you’ve done your homework. What is the system supposed to achieve, which machines are truly critical, and do you even have the data needed for analysis?
Five questions belong before the software selection: which goal is under the greatest economic pressure, which plants are business-critical, how good is the existing data situation, what infrastructure is still missing, and who will evaluate the data later? Data availability has a greater impact on success than the chosen platform.
| Question | Why It Matters Before You Buy |
|---|---|
| What is the goal? | Energy monitoring and early failure detection require different sensors and resolutions |
| What kind of equipment? | A pilot program involving three critical machines proves successful, while a blanket approach fails due to the sheer volume |
| What is the data situation? | Without a controller with a data output, the project begins with retrofitting, not with software |
| What kind of infrastructure? | Data storage, the network, and the gateway must be up and running; otherwise, the analysis remains purely theoretical. |
| Who does the evaluation? | Without a designated role, the system generates unread alerts after several weeks |
First, clarify the goal
Some companies want to detect downtime earlier, others want to reduce energy costs, and still others need documentation for customers or certifications. Each of these goals leads to a different selection of key performance indicators, sensors, and analyses. A system designed for energy monitoring rarely provides the high-resolution vibration data required for early failure detection. If you clarify your objective before selecting the software, you’ll save yourself the expense of a second purchase a year later.
In practice, these goals often overlap. Production management wants to avoid downtime, the finance department wants to reduce energy costs, and the management is asking for evidence for the next customer audit. Instead of tackling all three goals at once, a clear starting point helps. Which goal currently faces the greatest economic pressure, and where can success be demonstrated the fastest?
Which systems are truly critical?
The most common mistake during implementation is trying to monitor all machines at once, rather than starting with the business-critical systems that are most prone to failure (Deloitte, 2023). This approach is understandable, since more data feels like greater security. In practice, however, this leads to projects failing due to the sheer volume of data before any benefits become apparent.
It makes more sense to run a pilot program with equipment whose failure would immediately halt production or whose repair would be particularly expensive (Siemens and Deloitte, 2023). This could involve three machines or an entire production line. It is important that the pilot’s success can be demonstrated with concrete figures, such as fewer hours of downtime or faster response times in the event of malfunctions. The article on [...], discusses which thresholds are worthwhile for this. Predictive Maintenance.
How reliable is the available data?
According to a PwC study, data availability is the most important success factor for companies looking to implement equipment monitoring—more important than the chosen software platform (Siemens White Paper, 2018). Before selecting software, it is therefore worth taking a sober look at the current situation. Do the machines already have sensors and a control system that outputs data? Is there a network through which this data can be transmitted? Or is the project starting from scratch, with machines that have been running unchanged for twenty years?
These questions also determine the timeline. A facility with modern control technology can start seeing data within a few weeks. A facility with an established machine fleet first needs to be retrofitted, as described in the article on Retrofit Instead of Buying New .
What infrastructure is still missing?
Effective plant monitoring typically requires three components: a data storage system, a reliable network, and an IIoT gateway that converts machine data into a standardized format (Siemens, 2023). If any one of these components is missing, implementation will be delayed, regardless of how good the chosen software is. A common misconception is the assumption that the cloud platform alone solves the problem. Without a stable connection to the machine, any cloud-based analysis remains purely theoretical. This article explains how this chain works in detail From Sensor to Dashboard.
What is a realistic budget?
The software costs alone are only part of the equation at this stage. Added to this are sensors, installation, a gateway per machine or line (if necessary), and the time internal staff spend on implementation. Those who compare only the provider’s license costs often significantly underestimate the project’s total cost. A realistic budget can only be established once the installation and first twelve months of operation are factored in.
The issue of interfaces also needs to be factored into the budget. If the asset monitoring system is to interface with the existing ERP or maintenance system, additional integration costs will arise. If, on the other hand, the analytics are operated as a standalone tool, the effort is reduced, but the data remains separate from other systems. Both approaches are valid, but a decision should be made before selecting the software.
Who will analyze the data later?
During the selection phase, people often overlook who will be viewing the dashboards on a daily basis and who will respond to the alerts. A system without clearly defined responsibilities will result in unread notifications after just a few weeks. Maintenance management, shift supervisors, and the IT department should know before making a purchase who will assume which role in the new process. This organizational issue cannot be resolved retroactively via a software update.
Go through the list of questions togetherWe assess your current situation and candidly explain where it's still too early to implement software and where upgrading your existing system is the first step.
Frequently Asked Questions
How long does it take to implement equipment monitoring?
With modern control technology and a clearly defined pilot scope, the first usable data is often available after just a few weeks. With a machine fleet that has grown over the years and no sensors in place, preparation can take several months before any reliable data flows at all.
Cloud or on-premises solution—which is a better fit?
That depends on the company's IT infrastructure and security requirements. Cloud solutions scale more easily across multiple locations, while on-premises solutions are better suited for environments with strict data protection or network requirements. Both approaches require the three components mentioned: data storage, network, and gateway.
Who really needs the data in the end?
In most cases, multiple roles are involved simultaneously: maintenance for urgent alerts, production management for trends and planning, and, in some cases, quality management for documentation. Before implementation, determine which view each role needs, rather than giving everyone the same interface.
The Next Step
torck develops its plant monitoring solution as a standalone product and, based on its own projects in industry and retail, has the answers to questions regarding data availability, infrastructure, and budget. Our teams in Maxhütte-Haidhof, Vienna, and Rabat have already worked through this set of questions with companies that were in exactly this situation. You can find more information and schedule an initial consultation on the page for the Plant Monitoring.
German 
English 
French