Page 45 - Energize August 2022

TECHNICAL

APM Predict:
• Is easy to add into an existing operational
technology (OT) landscape, where it runs
as a native application on the edge, thus
providing connectivity to, for example,
ABB Extended Automation System 800xA
Publisher, ABB Symphony Plus publisher,
OPC UA servers, or Modbus TCP devices.
• Comes with prebuilt asset models
covering a wide range of assets which
address ABB offerings and target
industries. These models include assets
which range from simple sensors and field
instruments up to complex electrical, Figure 3: APM Predict: real-time condition monitoring on the edge
rotating and process equipment.
• Features an easy-to-use maintenance expand the basic models of APM Predict into first principle and ML models. In addition to these
workplace interface in which the user advanced modelling capabilities, it also provides an environment for clients to embed enterprise
can see condition notifications and knowledge and capture tribal knowledge.
recommendations, as well as organise One modelling concept which is gaining prominence is the digital twin. A digital twin provides
and understand assets across plant areas, an essential replica of a physical asset, system or process in digital form, enabling intervention
sites, or fleets of assets. before problems occur. Digital twins embody deep domain experience and apply physics-based
1
• Displays electrical assets in the electrical or AI/ML models to the behaviour they capture.
context via a single-line diagram viewer. It is within Predict 360 that digital twins are created to help predict failure and eliminate the
changes which happen in the “black boxes” into which operatives normally cannot see.
Asset condition information is provided Perform 360 epitomises the ability to gain greater asset insight by performing analytics on a
according to NAMUR NE107 (a field device combined OT and IT dataset. It integrates the contextualised data of plant IT systems, such as an
data standard), or according to an asset existing computerised maintenance management system (CMMS), with OT data streaming from
health severity scoring. It is possible to APM Predict modules. This provides plant reliability experts with a deeper event perspective
group and organise asset models, assign as well as a means to monetise different maintenance options, consider impacts of planned
a criticality to different task models and maintenance schedules and devise a means to extend intervention intervals and asset longevity.
provide a calculated view of the overall Perform 360 provides insights across asset performance, health, maintenance and life-cycle
health of an asset (Figure 3). costs. To simplify deployment, it is supplied with more than 40 out-of-the-box asset templates
Predict 360 extends the power of with prebuilt performance models and an array of embedded calculations (Figure 5).
predictive asset models from prebuilt It is also possible to assign different types of alarm and integrity operating window limits
versions to customisable versions specific and track these from dashboards in the Performance Monitoring Workplace. Compliance can
to an industrial operation. It provides also be tracked.
an environment for asset subject matter The value of Perform 360 lies in its use of preconfiguration to make effective deployment
experts and data scientists to continuously simple. The module leverages preconfigured key performance indicators (KPIs) with self-
capture and codify their knowledge into ABB
Ability Genix APM, which helps automate
diagnostic activities, reveal latent asset and
process issues, and discover opportunities
for higher production efficiency and asset
utilisation. Predict 360 helps in advanced
fault prediction, increasing uptime and
preventing failures of critical assets. The
module comes with a configurable asset
model library which has prebuilt dominant
failure modes. It is a self-service application
which allows any authorised person to set
and configure rules – or AI and machine
learning (ML) algorithms – to enable
predictive maintenance (Figure 4). It also has
a comprehensive workplace for asset fault
monitoring with recommendations.
Predict 360 provides users, such as
data scientists, with advanced modelling
capabilities which can be leveraged to Figure 4: Predict 360: Asset models definition illustration: Pump

energize | August 2022 | 43