Discovering new and exciting ways of improving data used for organisational decision making

Group lead: Dr Philip Woodall

It has been an exciting quarter for the Data Management Theme as the team presented the final results of the VIPr project to Boeing executives in September; the executives from Boeing Research and Technology visited Cambridge as part of their annual review of research at Cambridge. The VIPr project (Virtual Intelligent Production, Procurement and Prediction system) has developed a system to enable Boeing to make predictions about impending supplier parts delivery disruptions. The predictions allow the resulting parts shortages to be actively managed by giving advanced warnings to the teams that develop mitigation strategies within Boeing.

The team (including Pascal Wichmann, Alexandra Brintrup, Philip Woodall and Duncan McFarlane) presented the final architecture of the system illustrating how it can make predictions, despite the lack of supply chain data, by using combinations of publically available data and internal Boeing data. One of the key components of the architecture is the Supply Chain Miner, which can use publically available data to extract the supply chain of any company. This is a very exciting tool that the Boeing executives want to leverage immediately, and the team is in the process of transitioning this into Boeing. The executives also noted other purposes that the tool could be used for (these purposes are confidential to Boeing and so cannot be stated here) in addition to extraction of the supply chain, which is an unanticipated benefit of the tool.

Furthermore, the VIPr project has been extended until next August to allow further developments and extensions to be made to the various components of the architecture, including the Supply Chain Miner.

Visitor to the group

Tom Haegemans is visiting the group for 4 months to support the DASHLog project sponsored by YH Global Supply Chain Co., Ltd. Tom is a PhD candidate at the University of Leuven (Belgium) and is funded by a major Belgian financial institution. He also holds a Master’s degree in Information Management and a Bachelor’s degree in Applied Informatics. His research interests include: data quality measurement, representation of data quality measurements, causes of errors in manually acquired data, and data alignment strategies.

Figure: Aligning identical copies of data

Tom will work in cooperation with Dr Philip Woodall. Together, they will investigate issues related to alignment of two or more datasets (see the above figure) to ensure that information in one is properly updated in the other, and also to determine the optimal time to make any alignments so that up-to-date data is available for decision making. Interestingly, the master copy does not have to represent data in an information system: it could represent reality (such as the placement of physical items in a warehouse), and the slave copy represents the information system (such as a warehouse management system). In this warehouse case, the model could therefore represent when is the optimal time to perform an inventory check. Alternatively, it could represent a transactional information system which does a periodic batch transfer of its data to a data warehouse. The model, developed by Tom, is therefore very flexible to represent various cases.

Recent Publications

• Woodall, P. (2017). The Data Repurposing Challenge: New Pressures from Data Analytics. ACM Journal of Data and Information Quality, 8, ( 3-4 ), art. no. 11
• Gao, J., & Woodall, P. (2017). Topic Modelling in Information Quality Research – ICIQ 1996 to ICIQ 2016. International Conference on Information Quality (ICIQ).
• Giannikas, V., Woodall, P., McFarlane, D., & Lu, W. (2017). The impact of B2C commerce on traditional B2B warehousing. International Symposium on Logistics (ISL 2017).

These publications can downloaded from here. 

Publications in progress

(If you would like advanced access to any publications or research results, then please contact Dr Philip Woodall)

• Woodall et al. Potential Problem Tagging: Augmenting information systems with the capability to deal with inaccuracies.
• Wichmann et al. Who’s behind the curtain? Towards automatically generating supply chain maps from natural language text.
• Brintrup et al. Predicting hidden links in supply networks. (Submitted to Complexity Journal)