points that previously had no real-time process connection. Using DTs and the
associated peripherals, all relevant information in the value chain, from raw material
producer over the bioproduct manufacturer (e.g., food or pharmaceutics) to consumer, can be digitized. Concepts and strategies can be incorporated in the decisionmaking and quality assurance of bioprocesses.
Approaches from the field of artificial intelligence are a sophisticated method to
structure and evaluate the copious data. Sharing information from DTs supplements
the DTs with information from other DTs in the B2B sector. Intelligent algorithms
can use the information gained from the data for production planning and control.
The combination of established simulation and optimization methods as well as the
methods from the field of artificial intelligence and machine learning represents a
future-oriented technology. Besides, by processing data, KPIs, or forecasts and
events information may be included in the optimal configuration of system parameters. Therefore, predictions or behavioral analysis can be applied, and in some
cases, repairs of equipment (e.g., fermenters, pumps) can be ordered before the
downtime occurs. The digital strategy Predictive Maintenance reduces downtimes
and associated costs [87, 88]. These predictions are also applicable to resources.
Thus, detailed planning of raw material orders can be conducted, or energy can be
smartly distributed within the company or, if necessary, diverted to the respective
consumer. Furthermore, a detailed production layout planning can be implemented
with the help of DTs, where all process resources are used in a time-optimized
manner [89]. Hygiene concepts, hygiene state of the equipment, and the hygienic
monitoring of associated cleaning concepts can be monitored by DTMS. Moreover,
structured DTs in a DTMS help distribute the planning between production units and
production systems, improved decision support by simulation models, production
unit planning, and automated execution of offers and orders [4].
The comprehensive data collected and evaluated supports documentation of the
entire life cycle. This aspect enables higher traceability and new analyses on quality
assurance. In particular, the cross-linking, distant from rigid linear hierarchies,
creates higher transparency and, thus, new approaches to monitor and analyze
processes. A holistic view of the production and the company leads to new means
and methods of standardization, which lead to higher efficiency. Plant or product
data can be processed directly in the cloud and provided in real-time, and any
deviations are detected swiftly. Accordingly, one of the main challenges is to find
the root cause of the event within an acceptable time frame and without much effort.
Here, machine learning-based anomaly detection can be used. This approach does
not directly lead to a root cause but helps identify conspicuous time ranges and
production batches. The DTMS and cloud platforms enable patterns to be recognized
through various evaluation algorithms and, thus, processes and products can be
evaluated in detail and autonomously.
The Challenge of Implementing Digital Twins in Operating Value Chains
161
associated peripherals, all relevant information in the value chain, from raw material
producer over the bioproduct manufacturer (e.g., food or pharmaceutics) to consumer, can be digitized. Concepts and strategies can be incorporated in the decisionmaking and quality assurance of bioprocesses.
Approaches from the field of artificial intelligence are a sophisticated method to
structure and evaluate the copious data. Sharing information from DTs supplements
the DTs with information from other DTs in the B2B sector. Intelligent algorithms
can use the information gained from the data for production planning and control.
The combination of established simulation and optimization methods as well as the
methods from the field of artificial intelligence and machine learning represents a
future-oriented technology. Besides, by processing data, KPIs, or forecasts and
events information may be included in the optimal configuration of system parameters. Therefore, predictions or behavioral analysis can be applied, and in some
cases, repairs of equipment (e.g., fermenters, pumps) can be ordered before the
downtime occurs. The digital strategy Predictive Maintenance reduces downtimes
and associated costs [87, 88]. These predictions are also applicable to resources.
Thus, detailed planning of raw material orders can be conducted, or energy can be
smartly distributed within the company or, if necessary, diverted to the respective
consumer. Furthermore, a detailed production layout planning can be implemented
with the help of DTs, where all process resources are used in a time-optimized
manner [89]. Hygiene concepts, hygiene state of the equipment, and the hygienic
monitoring of associated cleaning concepts can be monitored by DTMS. Moreover,
structured DTs in a DTMS help distribute the planning between production units and
production systems, improved decision support by simulation models, production
unit planning, and automated execution of offers and orders [4].
The comprehensive data collected and evaluated supports documentation of the
entire life cycle. This aspect enables higher traceability and new analyses on quality
assurance. In particular, the cross-linking, distant from rigid linear hierarchies,
creates higher transparency and, thus, new approaches to monitor and analyze
processes. A holistic view of the production and the company leads to new means
and methods of standardization, which lead to higher efficiency. Plant or product
data can be processed directly in the cloud and provided in real-time, and any
deviations are detected swiftly. Accordingly, one of the main challenges is to find
the root cause of the event within an acceptable time frame and without much effort.
Here, machine learning-based anomaly detection can be used. This approach does
not directly lead to a root cause but helps identify conspicuous time ranges and
production batches. The DTMS and cloud platforms enable patterns to be recognized
through various evaluation algorithms and, thus, processes and products can be
evaluated in detail and autonomously.
The Challenge of Implementing Digital Twins in Operating Value Chains
161