3D Printing (3DP) has been regarded as a disruptive technology with myriads of benefits, this includes decentralized manufacturing, rapid prototyping, reorchestration and domestication of supply chains, lowering logistics costs, as well as sustainable and circular implications. However, limited focus has been given to integrating 3DP technologies with traditional manufacturing, particularly in the context of demand and production disruptions.

This study investigates the impact of utilizing a single 3DP as a production tool to mitigate disruptions, through considering production of 5 auto parts (side mirror cover, sun visor clip, door lock knob, rear view mirror cover, and a door handle), with a predefined set of demand and production size, the study assumes production, shortage and holding costs as percentages for traditional manufacturing, while calculating the actual production costs for 3DP, including two different infill rates.

On the other hand, to evaluate the effect of disruptions, a Monte Carlo simulation is employed to simulate different probabilities affecting the demand and supply quantities across a large set of instances. This allows to observe the effect of using a 3DP on different part sizes and through different scenarios of demand and production instances. The study uses MS Excel to perform various calculations.

Moreover, we assess the effect through considering several Key Performance Indicators (KPIs): demand fulfillment rate, holding cost, shortage cost, total production costs and unit production costs. Notably, the model shows that holding cost increased when combining with 3DP, as holding cost is a percentage of produced parts, suggesting that other manufacturing strategies as Make to Order, Engineer to Order and Just in Time are better fit for 3DP, thus decreasing holding costs and increasing demand fulfillment rates.

Additionally, part weight plays a major role as a variable, smaller parts mean more parts to be produced, increasing production and holding costs as well as demand fulfillment rate, while decreasing shortages. This is observed for instance for the sun visor clip, door lock knob, and door handle as demand fulfillment rates increase by 28%, 19%, and 15% respectively at a 20% infill rate. Conversely, as part weight increases, the benefits of 3DP regarding demand fulfillment rates diminish. Hence, reducing part weight through utilizing better 3DP designs approaches as Design for Additive Manufacturing (DfAM) practices would yield the positive benefits of higher fulfilment rate while minimizing production and holding costs.

This analysis also shows that 3DP can be beneficial under conditions of low production and high demand instances, with high shortage costs and low holding cost percentages. The practical implications of this study include providing stakeholders with a tool to assess the deployment of 3DP and study its effect on covering disruptions. While future development of the study and the tool can include change in 3DP costs, specifically labor costs, the consideration of other parts, different sets of demand and production quantities, and different production and inventory models. Moreover, validation based on practical case studies should be considered for more robust insights for stakeholders.

Introduction

Offshore wind energy is a pillar in the decarbonization strategy of the German government as there is enough space and steady winds in the North and Baltic Seas to make this a viable and important energy source. The logistics however pose a great challenge especially during the Operation & Maintenance (O&M) phase due to the great distances, harsh winds, and weather conditions at sea. AIS (Automated Identification System) data transmits every vessel’s position, speed, and course-over-ground, at certain intervals based on said speed, which can allow the routes of O&M vessels to be traced. The research was done within the LogReview project funded by the German Federal Ministry for Economic Affairs and Climate Action.

Methodology

To trace the routes of O&M vessels, a large AIS-dataset provided by the Danish Maritime Authority was used. This dataset consisted of 3.5 Terabytes of public available data, in addition to data collected in the LogReview project from three receivers based in the North Sea at the FINO 1, and FINO 3 research platforms, as well at the island of Heligoland. To extract meaningful data from this data set, the following steps had to be done: Data Pre-processing, Data Cleaning, Data Slicing and Data Analysis through means of the Panda library of Python. After these steps, the data was filtered for crew transfer vessels (CTV) and stored in a SQL Lite database. The coordinates in latitude and longtitude of offshore wind turbines, from the research platforms mentioned above, were stored in a library created for this project. The meaningful data was filtered into days and months with both statistical graphs and plotted tracks created using the Python library Folium. With the completion of these steps, it was possible to derive the exact tracks of the Offshore Wind O&M Vessels within the wind farms including the duration of their visit to a particular wind turbine. This further allowing the deduction of the time of O&M procedures.

Findings

The tracks of the Offshore O&M vessels based on AIS-data shows the O&M processes over a longer time period and can help to improve these processes. Better processes will lead to shorter distances travelled and thus lower costs which is one key part to make offshore wind more cost effective.

Discussion

By using AIS data, wind farm operators can use a new tool to make data driven decisions for their CTVs doing O&M. During the first few years of its lifespan, the original equipment manufacturer is responsible for O&M. But after the initial contract is settled, the responsibility shifts to the wind farm owner. Manufacturers typically are unwilling to share their O&M strategies, citing business secret protection. The analysis of AIS patterns will allow for identification of the optimal O&M schedule used by the manufacturer, to decrease costs for the wind farm owner when O&M ownership changes hands.

Conclusion

In conclusion this approach shows that AIS-data can be used to get a better insights into offshore wind O&M processes and provide an opportunity to improve these.

Additive manufacturing (AM) is rapidly advancing from prototyping to industrial serial production but faces persistent quality issues. Most current research occurs in lab settings, focusing on successful print quality and neglecting real-life data and errors that interrupt or spoil print jobs.

To address these shortcomings, we conduct a case study and conduct 13 interviews with AM service providers, AM manufacturers, and AM operators. We target companies in the transportation industry that have substantial experience with AM. This industry is particularly well-suited for AM adoption due to the long operational lifespan of vehicles and the relatively low installed base. With this study, we contribute to a better understanding of main AM printing challenges by asking (i) which factors lead to build failures? and (ii) how can firms develop learning curves to reduce build failures? We focus on the print job itself and excludes the preceding (e.g., part design, data preparation, or material supply) and subsequent (e.g., post-processing, quality assurance) processes. We use an exploratory research design building on qualitative interview data from various AM experts.

We find that build failures occur mainly due to three reasons. First, build failures often occur during the initial setup phase, typically within the first one to three print jobs, due to machine settings, part design, or material characteristics. While these failures are common in the ramp-up phase and must be planned for, they become rare in the serial production phase once print job repeatability is established, although software tools cannot eliminate them. Second, build failures often occur during the first few layers of a print job due to adhesion issues on the build plate, temperature or humidity issues with the material, or calibration of the extruder nozzle and its movement. Operators can mitigate these failures by closely observing the initial phase of the print job, thoroughly preparing and calibrating the printer, build plate, and nozzle, and storing materials in cabinets with controlled temperature and humidity as per ISO/ASTM DIS 52920 standards. Third, build failures can occur due to unforeseen external influences like power blackouts or physical movement of the printer, causing issues such as vibrations, clogged extruders, uneven layers, and extrusion stops. Although difficult and costly to mitigate completely, firms can plan for these events by installing backup energy supplies and positioning printers in locations with minimal physical impacts, though eliminating all possible external influences remains challenging.

We find that learning curves play a major role in mitigating build failures. If firms collect data about print jobs and build failures over time, they become capable of understanding the weaknesses of the print job and mitigating build failures. Learning curves can be developed over time, production volume, and knowledge of operators. Firms can achieve learning curves by focusing on single AM technologies and following quality standards (e.g. ISO/ASTM DIS 52920). Our findings contradict the image of using AM as “plug and play” and support the perspective that companies with little experience should consider using third party AM service providers.