Predictive Maintenance - Using Data from Building Technology to Reduce Malfunctions and Breakdowns

IT & OT

Predictive maintenance can effectively minimize failures and malfunctions and detect potential problems at an early stage. Sensor and system data as well as data analysis and machine learning are used to monitor and detect problems. Find out more about an exemplary implementation.

  • Predictive Maintenance

Predictive Maintenance

Less Disruption with Data from Building Technology

A well-known Swiss pharmaceutical company is creating a new network across the entire site to collect data on building technology, emphasizing the importance of an effective and predictive maintenance strategy. We assisted with the planning and implementation.

At a time when buildings are becoming increasingly complex and technology-intensive, developing effective maintenance strategies is crucial. The focus is on minimizing breakdowns and malfunctions. Predictive maintenance uses sensor and system data as well as data analysis and machine learning to continuously monitor the condition of building systems and detect potential problems at an early stage.

The implementation

The implementation took place in two project phases: In phase 1 , the network structure was set up first, then the systems were connected in phase 2.

The network structure consists of 35 industrial switches from Hirschmann's Greyhound series, which are operated via fiber optic cables in a ring network. In preparation, KRIKO planned the installation of new network technology and fiber optic connections for 41 existing network cabinets in 30 buildings. The core of the network was formed by a redundant firewall network from Fortinet for secure connection to the existing plant network. All incoming and outgoing connections are specifically released by the firewalls and checked for viruses, anomalies and unwanted attacks.

In the second phase, 195 network-compatible compact systems with a total of 213 network subscribers were connected to the newly created ring. These systems included UPS systems, battery monitoring systems, (emergency) lighting and shutter controls, as well as measurement systems.

Segmentation of the systems

The systems were already classified in the planning phase and logically segmented using VLANs based on the system types. The segmentation leads to a channeling of the data traffic with prioritization options, the avoidance of network overload and the increase of IT security.

KRIS as a data center

The data center of the entire network is a KRIS system from KRIKO, which collects, monitors and archives relevant data from a maintenance perspective and alerts maintenance personnel to critical system statuses. The connection to a cloud for data analysis using self-learning models is in progress.

The project was successfully implemented step by step over a period of almost two and a half years, from the conceptual design to the training of plant personnel, and serves as a guide for further optimization projects in the area of maintenance. Predictive maintenance optimizes the operation of buildings, reduces costs and increases user satisfaction by identifying and rectifying potential problems at an early stage.

Are you interested in an effective predictive maintenance strategy? We will be happy to support you with planning and implementation.
We look forward to hearing from you.


Go to the Contact Persons

Back