Was muss aiomatic über meine Daten wissen?

Wir benötigen Angaben dazu, welche Datenfelder überwacht, ignoriert oder als externe Einflussgrößen betrachtet werden sollen. Ebenso ist es essenziell, den Zeitpunkt zu definieren, an dem der Prozess den gewünschten Zustand erreicht und die Daten dem Normalzustand entsprechen.

Welche Arten von Daten kann die Software von aiomatic verarbeiten?

Unsere Software kann Zahlen, Dezimalzahlen, Texte sowie Wahr/Falsch-Aussagen verarbeiten. Dabei berücksichtig sie sowohl Sensordaten als auch das Fachwissen des Ingenieurs, das über eine benutzerfreundliche Oberfläche eingegeben wird.

Was benötige ich, um mit der Datenerfassung zu beginnen?

Im ersten Schritt ist die Bereitstellung einer virtuellen Maschine oder alternativ die Installation der zusätzlichen Hardware erforderlich (siehe Möglichkeiten der Datenverbindung). Im zweiten Schritt erfolgt der Verbindungsaufbau zum OPC UA-Server der Maschine, wofür gegebenenfalls Netzwerkanpassungen und Portfreigaben erforderlich sind.

Welche technischen Voraussetzungen gibt es bei der Datenverbindung über eine virtuelle Maschine?

Für die Ausführung unseres Datenverbindungs-Clients auf einer virtuellen Maschine ist eine Linux-VM und eine VPN-Verbindung erforderlich.

Project Procedure for Predictive Maintenance Solutions

This sets project work with aiomatic apart

Clear goals
& structured approach

Timed milestones and the provision of step-by-step instructions.

Comprehensive onboarding
& ongoing support

We guide your project to success through targeted workshops and regular meetings.

Highest security standards

Your data is stored in the ISO/IEC 27001-certified Microsoft Azure cloud and managed in accordance with the GDPR and the EU AI Act.

The 5 steps to success

Our project process at a glance

Phase 1: Kickoff & technical alignment

Together, we will define the final project parameters: which machines to prioritize, which data sources to use, and whether additional sensors are required. We will also align on goals, expectations, and the next steps.

Phase 2: Establishing the data connection (retrofitting with sensors may be required beforehand)

The data connection is set up using aiomatic’s standardised instructions.
Existing digital interfaces can usually be used without the need for additional hardware. If required, the data set can be supplemented with additional sensors.

Phase 3: AI Configuration

During this phase, we will work together to select the relevant data channels and configure the machine in the dashboard. The key machine areas and components are mapped in a way that makes them useful for analysis and monitoring. Based on this, the dashboard structure and AI training are set up.

Handover of the dashboard with digital machine structure.

The optimisation phase begins with the handover of the Digital Maintenance Assistant.

Phase 4: AI Model Optimization

The AI models are continuously trained and optimized with new data.
In joint bi-weekly meetings, we discuss detected anomalies, such as indications of wear or unfavorable operating conditions. This helps prevent damage and identify recurring patterns as well as harmful operating behaviors.

Introduction of the application: Scaling

A large part of the experience from the pilot plant can be transferred to structurally identical systems, which results in significant cost savings.

Thanks to the early warning from aiomatic, we were able to detect a gearbox failure in time. As a result, we were able to maintain operations over the weekend with reduced output and prepare for the necessary repair, instead of being surprised by a sudden imminent failure.

Ben Thurnwald, Managing Director
HANSAPORT Hafenbetriebsgesellschaft mbH

aiomatic is pioneering. For example, the application of AI to real-time data has already predicted imminent warehouse damage for one of our systems. As a result, we were able to act early and avoid an unplanned downtime of more than 8 hours.

Dirk Schlamann, Maintenance Lead
Nitto

The future of maintenance is difficult when it comes to skilled labor. With the AI-based solution from aiomatic, however, this looks much brighter. The team can now focus on other activities, while monitoring and early warning alerts are provided by aiomatic.

Andreas Weber, Technical maintenance manager
Canyon Bicycles GmbH

The biggest potential for you

Success factors for a quick & successful project start

For easy and fast implementation of our software and maximum added project value, we recommend these basics:

Machine data is already being digitally recorded periodically or on an event-based basis, at least every hour.

Machine data can be made available continuously via a digital interface.

IT staff from your team are available for the implementation.

Support

Frequently asked

Here you will find answers to the most common questions about the process of a Predictive Maintenance project.

What different solutions does aiomatic offer?

In addition to historical data analysis, we offer one-day workshops to plan your individual maintenance strategy. If you are unsure whether your data basis is sufficient, we would be happy to conduct a proof of concept with you so that you can test our software in your live environment. For long-term use of our software, our scalable SaaS model is the right choice for you.

What steps are required for data collection?

To collect data, you first define the relevant machine or process. The data can then be transmitted securely via encrypted MQTT or connected using our aioConnect solution. This requires a virtual machine, appropriate network permissions, and the relevant interface information, such as OPC UA. We will support the setup process with detailed instructions.

What personnel resources are required ?

For successful implementation, you will need a machine operator, a person responsible for maintenance/servicing, and an IT contact person for network shares. A contact person from production may be required to provide the OPC UA information.

What does aiomatic need to know about my data?

We need to know which data fields should be monitored, ignored, or treated as external variables. It is also essential to define the point at which the process reaches the desired state and the data corresponds to normal operating conditions.

How much effort is required on the customer side?

The effort required on the customer side is minimal, as aiomatic handles the majority of the implementation in the background. In one of our retrofit projects, for example, 20 use cases were implemented with only about 6 hours of effort from the customer.
This enables a quick and straightforward start without placing a heavy burden on internal teams.

Not sure where to start?

In our free machine and use case analysis,
we evaluate your potential for implementing Predictive Maintenance.

Schedule your analysis

Project process with aiomatic