The final project report from the feasibility study led by Dr Martin Baumers at the University of Nottingham is now online. “3D Printing Production Planning: Reactive manufacturing execution driving redistributed manufacturing” describes results of the work undertaken at the University of Nottingham in collaboration with HP Enterprise and Knight Graphics. The results of this project will be presented at the 3DP-RDM event 3D Printing Where and How on 31st January 2017 at the IfM in Cambridge.
Martin Baumers, Ender Özcan and Jason Atkin
3D Printing (3DP) technology, also known as Additive Manufacturing, is associated with significant potential for supply chain innovation by enabling manufacturing configurations delivering additional value through differentiated products, for example in terms of production location. It has been suggested that pursuing distributed 3DP supply chains may be the result of strategic deliberation, yet it is also frequently noted that 3DP is prone to higher unit costs than conventional manufacturing. In this report we summarise the findings of a project which establishes an understanding of the relationship between the commercial performance of 3DP and the characteristics of its operation through the development of a demonstrator system.
In most industries 3DP faces the additional challenge of integrating with conventional manufacturing technologies and processes, which are normally operated in a centralised location. Economies of scale form one of the reasons for the dominance of centralised manufacturing, allowing the amortization of substantial costs over large volumes of products for the global marketplace. Complementing such centralised manufacturing, increasingly complex supply chains have emerged. Moreover, the implementation of appropriate supply chains is now seen as a core capability for manufacturing businesses.
In this context, a closer inspection of the typical work flow of 3DP reveals a puzzle: the current process for allocating build requirements to individual (potentially re-distributed) 3DP systems relies on an array of decisions on the operator/technician level, some of which can be automated and all of which will affect the overall efficiency of the 3DP process and the business case for its application. Thus, this project focusses on designing a demonstrator to provide a solution to this puzzle and further our understanding of the implementation of 3DP supply chains. Labelled the “3D Packing Research Application Tool” (3DPackRAT), the software demonstrator shows an avenue to the release of significant additional value by enabling a truly flexible and reactive manufacturing execution methodology that complements the strengths of 3DP. As an integrated approach to the build volume packing and scheduling problems encountered, this tool aims to determine the most appropriate 3DP system for each individual order in an automated process.
Effectively, such tools may allow adopters to leapfrog the gradual evolution of supply chain practise in response to the emergence of 3DP. To achieve this, the integrated computational framework we have implemented enables the inclusion of a wide range of general and location-related aspects in a single optimisation-based production planning procedure. Being fed an order stream, 3DPackRAT is designed to determine the best 3DP system for each build request. Crucially, this approach is also capable of considering the benefits resulting from re-distributed 3DP, driving supply chain structures towards such configurations where beneficial.
Our project was carried out as a collaborative and interdisciplinary programme of work between the 3D Printing Research Group (3DPRG) at the Faculty of Engineering and the Automated Scheduling, Optimisation and Planning (ASAP) Research Group at the School of Computer Science, both at the University of Nottingham, over the course of 2016. The programme consisted of three elements:
- Formulation of a portfolio of algorithms, heuristics and operational policies capable of addressing the combined build volume packing and scheduling problem for the baseline 3DP technology (Laser Sintering);
- Development and implementation of the demonstrator system 3DPackRAT up to the status of a minimum viable product, allowing experimentation by 3rd parties;
- Validation of the demonstrator system by interaction with a group of four industrial domain experts through presentations and live demonstrations, among them high-calibre international members of industry.
The project has demonstrated the possibility of implementing a joined-up approach to build volume packing and machine scheduling in 3DP and has shown that the problem in reality needs to be addressed in a multi-machine and multi-time period setting. Moreover, we have extended existing computational build volume packing methodologies to incorporate temporal aspects to address this problem.
Our interactions with various industrial partners advising the project indicate the requirement for such functionality. The demonstrator described in this report thus provides a first step in addressing the emerging workflow optimisation problem in 3DP in an integrated way.
[Image source: Martin Baumers]