Question 1

What exactly does ai-omatic solutions do?

Accepted Answer

ai-omatic offers a software as a service solution in the field of predictive maintenance. The basis is an innovative method developed by ai-omatic, which can use a complex mathematical model to detect and evaluate anomalies without previously defined limit values. The goal that ai-omatic is pursuing with this software is unattended, fully automatic monitoring that lives up to the term predictive maintenance.

Question 2

What is the flow from unstructured data to predictive maintenance?

Accepted Answer

The first step in the process is a workshop with ai-omatic, where the specific use case is discussed and goals for the project are set. This is followed by a quality check of the existing data to determine whether it is suitable for the use case. If data is not yet available or does not meet the quality requirements, data must first be collected. With these, ai-omatic can then train a model. Next comes the two-month proof of concept, in which the basic feasibility of the project is proven in the application. After that, the process transitions into a monthly licensing model. The status of the machines can be viewed continuously via the user-friendly dashboard, failures can be predicted and maintenance costs can be reduced.

Question 3

In which regions can the software be applied?

Accepted Answer

The software is not limited to any particular region or country.

Question 4

Who is the software suitable for?

Accepted Answer

The software from ai-omatic is basically aimed at all companies that have an interest in preventing unplanned machine downtimes. The software is particularly interesting for companies where failures lead to major cost and productivity losses The primary target group are machine manufacturers and production companies that use machines whose complex systems produce large amounts of data and whose regular operation is constant over long periods of time.

Question 5

For which machines is the software suitable?

Accepted Answer

The ai-omatic software is suitable for all machines that have mechanical motors or that rotate and whose monitoring by sensors produces large amounts of data. For example, ai-omatic works with machines from the automotive, shipping and manufacturing sectors.

Question 6

Which sensors are needed?

Accepted Answer

There must be sensors that produce large amounts of data and can give information about the problems that are occurring with the machine. In mechanical systems, these are mostly vibration sensors. The more context information is available, the better the software can work. Therefore, in principle, the following applies: the more sensors and data, the better.

Question 7

What can I do if I don't have any sensors yet?

Accepted Answer

Companies whose machines do not have suitable data-producing sensors can retrofit the required sensors. ai-omatic can support companies with such a retrofit.

Question 8

With what frequency must data be collected?

Accepted Answer

ai-omatic requires data that is twice the frequency of what is to be monitored. Using an example, this means: a sampling rate of 2kHz is required for a motor that rotates at 1000 revolutions.  Because the required frequency depends heavily on the specific application, it is difficult to define guidelines. Based on experience, however, it can be said that there must be at least one measurement per second (1Hz) so that the data is suitable for the software. High-frequency data is ideal, i.e. data in the 1kHz range and more.

Question 9

In what format does ai-omatic need the historical data?

Accepted Answer

Historical data can be transferred in different formats, a format that is often used is CSV.

Question 10

From which period should the historical data set be?

Accepted Answer

In general, at least one month of historical data should be available. However, the amount of data required is highly dependent on the frequency of the data and the type of machine. For machines with very constant, homogeneous behavior in normal operation, which generate data in the high-frequency range, data of less than one month is usually sufficient. For non-high-frequency data in the Hertz range, this period increases to around 1-3 months.

Question 11

If no historical data is available, how long does it take before a model can be trained?

Accepted Answer

Again, the amount of data required depends on the frequency and type of machine. In general: ai-omatic usually needs less than a month for data in the kHz range, and about 1-3 months for data in the Hz range. a511-2eb009ed8a2d>

Question 12

How does ai-omatic access the data?

Accepted Answer

If historical data is available, the data is sent to ai-omatic as a CSV file. If there is no data yet and it has to be collected live, ai-omatic can access the data via the interfaces OPC UA, MQTT, Modbus TCP or REST API (in JSON format).

Question 13

How much internet bandwidth do I need for the software to work?

Accepted Answer

The bandwidth required for sensor data streams depends on the amount and frequency of the data. In most cases, a standard Internet connection of 50 Mbit/s is sufficient.

Question 14

Where is the software running?

Accepted Answer

The software runs in the Azure cloud. In selected cases, ai-omatic also offers a more complex and cost-intensive on-premises solution, in which the software is integrated into the customer's infrastructure -2eb009ed8a2d>

Question 15

How does ai-omatic process my data and how is it used?

Accepted Answer

ai-omatic takes the issue of data protection very seriously. For this reason, data is only used as part of the project work with the customer and not beyond that. From January 2022, ai-omatic is also certified with the international standard ISO/IEC 27001, which specifies requirements for the establishment, implementation, maintenance and continuous improvement of a documented information security management system, taking into account the context of an organization and requirements for the assessment and treatment of information risks defined.

Question 16

What are the costs?

Accepted Answer

ai-omatic's aim is to enable companies to get started with predictive maintenance as easily and cost-effectively as possible. For the workshop, in which the use case and the goal are discussed, costs amounting to €1500, the quality check of the data costs €500 and a one-time setup fee of €15,000 is charged for the proof of concept. In the subsequent SaaS model, the price depends on the desired model: the essential model is €499 per month, the premium model is €699 per month and the custom model starts at €899 per month.

Question 17

What exactly do I finally get from ai-omatic?

Accepted Answer

The customer gets a user-friendly dashboard, which gives an overview of the vitality status of the machines at first glance. This allows failures to be predicted up to six hours in advance, thereby drastically reducing maintenance costs. New machines or data sources can be added easily and intuitively by the customer and produce new results within minutes. Thanks to the web-based software, access is possible from anywhere.