Key Attributes for Supply Chain Modeling
Today we have Part 2 of Lori Gordon's series on supply chain risk. Read Part 1 here.
The global pandemic recently cast economic shock waves throughout the globe. Looking deeper we see it revealed latent, underlying, systemic challenges in our approach to analyze and make decisions on data.
One of these challenges was China, a significant source of U.S. supply and the source of more than 40% of the world’s personal protection equipment (PPE) imports. China shut down PPE production to secure their own system, quickly immobilizing public health sectors across the globe and putting people around the world at severe risk.??
But the challenge was not unique to China. The Journal of General Internal Medicine, for example, found that the three largest PPE manufacturers — 3M, Honeywell International, and MSA Safety — did not disclose basic supply chain information, which contained errors in locations of manufacturing facilities, production levels, and dependence on foreign countries [1, 2]. Not only did public health officials in the U.S. and the UK have little visibility into their suppliers, they couldn’t trust the data that they were getting. It became necessary to acquire from new, unvetted sources.??
What was even more challenging was that this was not just limited to the public health community. Every critical sector — from energy to manufacturing to financial — suffered from similar issues. [3, 4, 5]?
The space enterprise — from national security to civil to commercial space — is one of the sectors still facing challenges [6].
To better understand solutions to mitigate longer term Coronavirus (COVID-19) effects on supply chains, the 5th Annual Innovation Summit hosted by The Aerospace Corporation in the fall of 2020 discussed these challenges within the Federally Funded Research & Development Center (FFRDC) and University Affiliated Research Center (UARC) community. The discussion centered on the need to develop a resilient supply chain in the face of a world pandemic. The group teed up an array of strategic national policy-level recommendations which included:
To support these strategic recommendations, organizations need greater insight into the supply chain for business decisions [7].
Illumination models can integrate and help analyze this information, providing context to make better, more risk-informed decisions through visualization of data that would not otherwise be intuitive. But illumination models are only as good as their data. What are some critical data attributes to ensuring good illumination models?
Time Sensitivity?
Leveraging real time data is important when data quickly changes in time. During COVID-19 businesses were undergoing significant changes in days and weeks, not months or years. Commercial aviation layoffs and plant closures were accelerating, and quarterly financials were unreliable and didn’t reflect lagging effects. Technology scouting for tools (AI/ML) are being used to harvest other types of signals — ?to mine the enormous amounts of unstructured data “hay” — to identify insights or trends that would indicate a potential risk or instability in a supplier.
Trust?
Trusted data is critical — garbage in, garbage out. While getting to 100% trust is challenging, we can improve trust by: selecting sources that are appropriate, certified, and diverse; processing data to transform and aggregate it; and collaborating cross-functionally to govern data for compliance. This will help eliminate issues down the road with quality, provisioning, and the risks of counterfeit products.?
领英推荐
Visibility?
Often the data available is reliable but incomplete. Partial data sets can serve as indicators but may be insufficient to provide a “big picture” view of real-world issues and trends. In many cases this leads to the “project by project” mentality. Investing in the identification of data that needs to be reported is crucial.
Relationships
Having a broad set of data to query can help identify relationships that are not immediately apparent. For example, a national database query of suppliers who produce valves might reveal that of the five U.S. companies that supply valves, four supply the Ford Motor Company. If Ford has high demand this might result in a squeeze to other industries, including the space sector. Mining this data can also help answer dependency and interdependency questions such as “If Valve Company X is unable to supply Ford Motor Company with X product, will they similarly not be able to supply the space industry?”
Conclusion?
The global pandemic challenged supply chains and our ability to analyze and make procurement and acquisition decisions. Key attributes necessary when modeling supplier and product marketplace data include time sensitivity, trust, visibility, and relational factors.??
Stay tuned for a forthcoming post which will discuss illumination models that provide a framework to make better, more risk-informed decisions through visualization of data that would not otherwise be intuitive.??
References?
7.?????https://democracycollaborative.org/sites/default/files/2020-11/HAN%20Reimagining%20PPE_web_1.pdf?
Lori W. Gordon leads space enterprise integration initiatives in the Corporate Chief Engineer’s Office at The Aerospace Corporation.
Getting It Right?focuses on industry collaboration for mission success by sharing lessons learned, best practices, and engineering advances in response to the nation’s toughest challenges. It is published by the Aerospace Corporate Chief Engineer's Office.