The overall objective of intelligent maintenance is thus to shorten the latency periods (delay times) from when an event occurs until the appropriate remedial action is taken by using technical digital aids, and thereby to increase efficiency. Three main areas of action were identified, and the individual tools can subsequently be assigned to them.
It is not possible to describe all maintenance activities for special-purpose machines in work instructions, since very few of the tasks are standard activities. Each repair must be considered individually on the basis of how the damage occurred and the damage pattern. This means that it is not possible to provide a generally applicable instruction that must be followed. The individual knowledge accumulated in the form of the maintenance engineer's experience is a key contributor to efficiency at this point. This experience is only passed on to younger colleagues verbally, if at all. Knowledge management was introduced to prevent precisely this loss of knowledge, which is caused by staff turnover and experienced employees retiring. Experience is documented and made accessible to colleagues in so-called "WiDoks" ("Wissensdokumente" or knowledge documents). A WiDok describes a procedure for solving a problem that has been used successfully in the past.
Maintenance staff are rarely based at an office workstation, but travel around the plant site to perform the necessary maintenance tasks. Access to information of any kind at any time via mobile devices such as cell phones or tablets has become indispensable in people's personal lives and in the IT sector. This is why all maintenance staff should be able to switch to mobile devices if they want to see their open orders in the SAP, need a component description, or want to know whether a spare part is in stock and where it is located. The development of a "smart maintenance platform" in Sharepoint, combined with a tablet, means that the required information can be accessed on site at the workplace.
Smart glasses are used in the AMAG rolling mobile maintenance system to provide (internal or external) specialist support to colleagues on site. These "head-mounted tablets" can be integrated into virtual meetings via the tried-and-tested "Microsoft Teams" software, thereby allowing all participants to see the situation at the plant for themselves and to provide assistance as necessary.
Condition monitoring / Predictive maintenance
Continuous condition monitoring is installed on most of the main drive trains of the rolling mills (roll and coiler drives) in the form of vibration monitoring with acceleration sensors. They calculate the damage level of the installed rolling bearings and gear teeth on a continual basis, thus enabling timely repair without risking gearbox failure. Another data source that indicates the technical status of a plant is the alarm messages programmed in the plant control system, which are displayed to the plant operators via the HMI (human machine interface). Using the in-house programming of our "smart message analysis", all fault messages of all production plants are collected in a common database and evaluated in such an intelligent way that we are able to focus on incipient defects.
The predictive approach, as a continuation of condition monitoring, represents the peak of maintenance strategies. Predictive models, which are usually based on AI or machine learning, also require a minimum set of data sets that reflect the component failure to be predicted for their training phase.
Drones as an indispensable part of inspections
Drones have become indispensable for AMAG facility management, plant maintenance, and the AMAG plant fire department. For example, drones are used to carry out necessary inspections of facility rooftops as part of building maintenance as well as green zones and traffic areas, such as the nearly 5-km-long stretch of AMAG's proprietary connecting railroad. The AMAG plant fire department uses a device with an integrated thermal imaging camera that is so sensitive that it can even detect game such as pheasants or fawns in AMAG's green spaces.
Facility inspection by means of remotely controlled camera vehicles
The maintenance of furnace bottoms on smelting furnaces is a good example of facility inspection scenarios. The welds at the bottom of the furnace are a possible weak spot in smelting furnaces. Welds can be damaged over time due to corrosion or expansion of the furnace housing, leading to leakage of liquid metal in the worst case.
Inspecting the welds at the bottom of the furnace is therefore extremely important in order to detect possible damage at an early stage and take appropriate measures. Compared to conventional inspections, unmanned inspection offers several advantages: speed, cost savings and, above all, huge improvements in occupational safety, as even dangerous or difficult-to-reach areas can be inspected at any time.
For all the digital tools described above to be successfully introduced and sustainably operated, we need not only a motivated and receptive maintenance team to generate ideas and move them forward, but also the support of the IT department and data scientists. Without this interdisciplinary collaboration, it will not be possible to further develop and expand the use of smart maintenance in the future.
Would you like to know more?