Figure 1: Drones can enable safe access, even to the inside of equipment

(Photo © Flyability)

Reliable data infrastructure 

Data from many different sources such as process, machine, laboratory and of course maintenance data is currently digitised and interconnected for automated and efficient processing. A first and relatively simple way to increase efficiency through digitalisation is therefore to implement workflows with a high degree of automation based on this data. This may sound simple, but providing data from different sources, such as sensor data, spare part information or data on the availability of staff is just as difficult as ensuring connectivity across the whole plant. Powerful digital support tools only deliver their added value if the data infrastructure is available and reliable enough.

Preconditions for application

Although we talk about “digital tools”, these cannot just be simply plugged in, but need to be carefully integrated into the plant’s environment. The correct and sustainable storage of data is one of the key challenges of digitalisation. Digital tools providing new data need to be connected to one side of a data storage solution, whilst data-based tools need to be connected to the other side. But what is the right way to store data in between? What information is required besides an actual measurement and its timestamp? There is a need for machine-readable meta-data containing information on the data quality, the plant and equipment structure, the type of sensors and their calibration and ranges. There are many solutions for the gathering and evaluation of data. But many of these “black boxes” do not provide any insight into the data itself. Ownership of data and the possibility to share data efficiently through standardised interfaces is essential for the application of different digital tools in cement plants.

Data quality cannot however only be ensured by standards and storage. Required sensors need to be installed, calibrated and maintained. Alarms need to be set according to the specifications of the installed equipment, and they need to be updated anytime the equipment is changed. Finally, database information such as stock availability of spare parts also needs to be maintained. This requires time and effort. All the ‘brave new world’ benefits therefore come with a price to pay.

Data-driven decisions

Technically embedding tools into the plant’s data structures, ensuring data quality and directing alarms into the central control room or into the offices of the maintenance department is only half of the journey. The big question is how the provided data, and in particular alarms are handled from this point onwards. Ultimately, the software cannot fix the mechanical problem.

An example: A bearing is monitored with sensors for temperature and vibration. Both parameters increase slowly but steadily. An alarm is displayed on the CCR screen, on the maintenance manager’s tablet or an outsourced contractor’s. Will the alarm simply just be acknowledged? Who is responsible and who has which authority when such an alarm occurs? Will someone inspect the equipment? And will the relevant part be exchanged prior to a possible failure? There are no simple answers to these questions. It depends on the priority and system relevance of the machine. Furthermore, the availability of qualified maintenance staff (Fig. 2), the special tools required, spare parts and the downtime due to repair need to be taken into account.