Page 135 - Kaleidoscope Academic Conference Proceedings 2020
P. 135

Industry-driven digital transformation





























                                            Figure 3 – Digital platform user interface



           platform uses different types of message brokers, i.e.  in the case of the geofencing use case, the machine positions
           RabbitMQ [32] and Kafka [33].                      are visualized in the main view component on a map, where
           Afterwards, the data is forwarded to the data lake, which  the user can define geofences and assign machines to these.
           consists of hot and cold storage for short-term, as well as  In case a non-allowed machine enters a geofence, the user
           long-term, savings of the data respectively, i.e. Cassandra  receives a warning notification and the incident is saved in a
           [34] and mongoDB [35]. Some data, i.e. machine positions  geofence incident report.
           can be directly sent to web clients using WebSockets.
                                                                      4.  CONCLUSION AND OUTLOOK
           Additionally, there is the data analysis component, which
           consists of a set of services for analyzing the data supported
                                                              4.1 Conclusion
           by machine learning approaches. This component aims at
           providing different utilities from historical data analysis to  In this paper, we presented a data-driven approach for the
           live monitoring. Moreover, it allows a behavior analysis of  digitalization of construction sites. For this, the requirements
           machines in comparison to simulation models and digital  of today’s settings have been analyzed and considered in
           twins.                                             order to address the existing gap of interoperability between
           The job controller is at the core of the communication  different actors on current construction sites. Moreover, our
           infrastructure. This module aims to manage different queries  approach helps to increase the productivity and efficiency
           and application programming interface (API) calls between  in construction processes thanks to various functionalities,
           all the other components of the infrastructure.    such as live monitoring or process analysis.  Certainly,
           Finally, the digital process platform represents the main  data privacy, data governance, and the security of machine
           interface between the user and the system. The platform  networking constitute an important topic for both machine
           enables the user to interact with real-world use cases of  manufacturers and platform users.
           construction sites.  The whole off-site side is hosted on  We agree with the examined literature that building digital
           Kubernetes, an open-source container-orchestration tool for  infrastructures are essential for many industry sectors to
           deploying, managing, and scaling of containers and services.  thrive.  Here, our current approach can be extended to
           A container is a software package that encapsulates all its  other industrial applications, such as farming machines in
           dependencies and can be deployed in different environments.  agriculture. Our proposed digital process platform shows a
           Here, the suggested platform relies on Docker containers.  first approach towards a digital construction site by providing
           Figure 3 shows a mock-up of the user interface in the digital  the necessary data processing pipeline from data acquisition
                                                              over data transfer to data integration and visualization.
           process platform. In the top navigation bar, represented in
           green (1), the user can check messages from other users, and
                                                              4.2 Outlook
           notifications from the system. The user can also access his
           profile, set his preferences, and log out when needed. In
                                                              For the future, one major goal is the integrationof digital twins
           the left menu bar, represented in blue (2), the user has an
                                                              for construction machines into the digital process platform.
           overview of the supported use cases, as well as the individual
                                                              In this regard, we plan to develop simulation models that
           component in the system.
                                                              make it possible to generate additional data on productivity,
           The main user interaction interface is represented in yellow
                                                              efficiency, and machine condition adding to the sensor data
           (3). Here, each use case has a predefined view that provides
                                                              of the real machines, which cannot be measured on the
           the user with the corresponding functionalities. For instance,

                                                           – 77 –
   130   131   132   133   134   135   136   137   138   139   140