The Fraunhofer Center for Maritime Logistics and Services (CML) carried out a comprehensive energy‑monitoring project in the Port of Brake, in cooperation with the Institute for Computer Science (OFFIS), the shipping company J. Müller AG, and the infrastructure provider NPorts. The project, funded under the German research grant FKZ 19H19005C, ran from early 2019 through late 2020, with a final report issued in 2021. The main objective was to create a real‑time measurement and data‑analysis framework that would enable the port to identify energy‑saving opportunities and to model future consumption patterns.
The technical work began with a clustering of consumer groups based on historical consumption data and process mapping. The resulting clusters formed the basis for the “TAP 2.4” infrastructure concept, which defined roughly 500 measurement points across the port’s electrical consumers and distribution boards. Each point was assigned a unique identifier and placed within a hierarchical structure (feed, sub‑distribution, individual consumer) to allow later attribution of measured values to specific processes. The infrastructure concept was jointly developed by CML, OFFIS, and the industrial partners and served as the blueprint for the subsequent installation of smart meters.
Installation of the smart meters was carried out by the industrial partners, who then transmitted the data to OFFIS via a secure VPN connection. The data stream was enriched with secondary information from ship lists supplied by J. Müller AG as HTML pages. Despite COVID‑19 restrictions that delayed on‑site work, the team accelerated front‑end and UI development for the dashPORT dashboard, ensuring that the data could be visualised and analysed in near real time.
Data analysis proceeded in several stages. First, an exploratory analysis of a three‑month subset of the collected data revealed errors and gaps, which were corrected in close collaboration with the partners. A descriptive statistical analysis then established typical consumption patterns and defined “off‑thresholds” – the minimal consumption of a switched‑off consumer that represents the lower bound of optimisation potential. These thresholds were validated with operational experts from the port.
A key contribution was the construction of a process‑dependency matrix. Using screenshots of the port’s process control system, the team manually traced the upstream and downstream relationships of each consumer. The resulting vectors of dependent consumers were documented in tables and validated with J. Müller AG. This dependency information is essential for generating realistic shutdown recommendations that do not disrupt critical process chains.
To improve forecast accuracy, the project re‑activated an old AIS antenna at CML to capture berth occupancy data. Geofences were created for each berth, and a decoding program translated AIS messages into a universal text format. The occupancy data were merged with the consumption records, allowing the training of a machine‑learning model by OFFIS that predicts future energy demand based on berth utilisation. This enriched dataset also enabled the identification of synergies across the entire port, as the electrical network was originally distributed through a central ring system managed by NPorts.
Throughout the project, weekly meetings between CML and OFFIS ensured alignment of data handling, modelling, and dashboard development. The final deliverable, the dashPORT prototype, integrates real‑time consumption data, process dependencies, and forecast outputs, providing port operators with actionable insights for energy efficiency and flexibility measures.
In summary, the project delivered a fully integrated measurement and analytics platform for the Port of Brake, comprising 500 smart‑metered points, a validated off‑threshold catalogue, a process‑dependency matrix, and a machine‑learning forecast model that incorporates berth occupancy. The collaboration between Fraunhofer CML, OFFIS, J. Müller AG, and NPorts, supported by the German research grant FKZ 19H19005C, demonstrates a successful partnership model for maritime energy optimisation.