A hurdle race to the OEE data

Hardware development is a complex undertaking. What applies to established companies such as Apple, is only right for start-ups validity. If a start-up with little previous experience, a small team and a minimum budget starts this adventure, the team needs good nerves. A hardware development process can lead directly into disaster. Why did we still risk it?

oee.ai’s business model requires capturing data from machines in factories. To get this data it’s an intensive physical and argumentative hurdles with the responsible discussion partners, which must be mastered again with each customer. But a honey pot attracts: The market of manufacturing companies with equipment use is huge. The language of the data is international. This puts us in front of a multi-billion dollar market that remains closed until someone masters this challenge – or find a way without hurdles.

6 hurdles on the way to capturing OEE data

What are the hurdles we are talking about? The investments in companies are sometimes so old that they themselves do not collect any data. We are in a market for capital goods. 10 years is considered as the lower end of the lifespan in a large part of the industries. Not infrequently, equipments are also used for 30 or more years. And no one disposes a functioning machine, only to buy a similar one with more sensors and a controller. The topic “Retrofit” will accompany the 4th Industrial Revolution for a long time.

Image: Very old machine, no longer suitable for oee.ai

Even when data is collected in the machine, it is not necessarily in the right place in the process or complete for our application case. For example, we need the interaction of quantities and loss reasons that are assigned by the employee via a display from a fault catalog. Such a functionality has no equipment that has not been specially designed for it.

If a modern controller is available, it is very reluctantly modified for two reasons. Firstly, especially medium-sized companies often lack employees with the corresponding know-how. Programmers for programmable logic controllers (PLC) are scarce. Even many mechanical engineering OEMs rely on Freelancer. Those who have employees with this expertise in their own ranks, can be lucky – but this is more the exception than the rule.

As a second reason, one fears a risk over the modification – “never change a running system”. These are production plants that have planned their resources for the long term. The explanation “we wanted to try something” is not a suitable explanation by a production manager to his COO for an unplanned shutdown.

However, the hurdles have not yet been completed. If the systems are modern enough, the system manufacturer can typically remunerate the access to the data with a separate license key. Thus, not infrequently five-digit amounts due in order to access their own data. In this case, usually additional software is needed such as an OPC UA server, which ultimately outputs the necessary format of the data – of course connected with additional costs and corresponding IT know-how.

And then there is the vague fear of connecting the systems to the Internet. The Stuxnet computer worm from the year 2010 has anchored itself in the collective memory. The malware was specially developed to attack a system for monitoring and control (SCADA system) from the manufacturer Siemens – the Simatic S7. Confidence in your own IT department or understanding of (non-IT) management of the operation of firewalls is limited. That’s how anxiety triumphs and valuable asset data is neither recorded nor evaluated.

OEE data acquisition without system intervention

An IoT gateway that captures data into the system without any system intervention or return channel must therefore be developed for our business model. We’re basically hardware developers against our will. But the best artificial intelligence and machine learning algorithms bring nothing if they are cut off from the data stream. We have this IoT gateway now. Developed in Aachen, produced in Shenzhen. Or as Apple would say: Designed in North Rhine Westphalia, assembled in China.

Incidentally, we are convinced that all the above arguments are outdated in a few years. In the future, it will become much easier, if not standard, for plant data to be accessible at a central location and in a uniform format. But we want the data now and until the future becomes reality, we provide our customers with our own, minimally invasive IoT gateway – as hardware-as-a-service for rent.