How Machine Learning Solves the COVID-19 Supplier Discovery Dilemma
Machine learning, in addition to a large amount of data, is a critical tool that can help manufacturers and supply chain professionals find, qualify, and use new suppliers for personal protective equipment (PPE) and raw materials during the COVID-19 pandemic. As stated in a recent Inbound Logistics article, “Companies need to first quickly develop transparency in their supply chains. This can be done by starting to map the location of first-, second-, and third-tier suppliers to know who is supplying what from where.”
PPE is currently one of the most in-demand categories of product worldwide, as consumer groups expand to include first responders, federal and state governments, grocery workers, retail location staff, and even the general public. It includes face masks, goggles, disposable gloves, and protective gowns. Sourcing N95 masks, for example, may seen relatively simple, but the mask actually requires eight layers: Spun-bond polypropylene and hydrophilic plastic, cellulose with copper and zinc ions, melt-blown polypropylene, more spun bound polypropylene, aluminum, steel, spandex, and polyurethane. It’s actually the polypropylene which makes the mask N95 grade (removing 95% of air particles because of the 0.1-micron holes in the material which block any particles of a larger size). So, you can see that this is a complicated task, that requires harnessing a diverse array of suppliers.
Companies in all industries are struggling to ensure they have the PPE on hand to continue operating, and that is no easy feat. We know that some products and materials are scarce, and visibility into the supply chain is limited. The problem is compounded by that fact that a product may not be available at the usual wholesaler or distributor, or there may be stock in a regional warehouse or somewhere mid-shipment, which can’t be located or purchased. In the midst of a disruption such as the one we are facing today, what matters most is speed, agility, and risk mitigation. Fortunately, machine learning can deliver on all three of these important factors.
Speed and Scope
Procurement solutions with machine learning capabilities allow companies to search global supplier information far more quickly and comprehensively than traditional means such as standard Internet searches, static category listings, and traditional supplier databases. Since the parameters for each search can be clearly defined, technology is a far better supplier discovery mechanism than a human resource following a cumbersome and time-consuming manual process. Powering your vendor master with machine learning and analytics will provide a unified, transparent and dynamic view of every supplier you work with. It will help you gain instant access to your entire supplier master with deep insight into supplier capabilities and relationships with your organization. In addition, you will be able to filter suppliers by spend, relevance, diversity status and geography.
In addition to improving the search process, machine learning can be leveraged to ensure supplier profiles remain automatically up to date, regularly crawling websites to pull in the latest information about supplier capabilities. And no matter what terminology is used on those websites, machine learning can standardize profile listings, further streamlining the search process and expanding search results.
Clarifying with Context
Finding a new or alternate PPE supplier is one thing. Pre-qualifying each source, deciding which to contact for more information, and selecting suppliers for a contract is another effort entirely. Companies still need to be purposeful about the suppliers they partner with, even in a time when desired PPE supplies are critical and scarce. Machine learning can incorporate important factors such as peer reviews and also search for specific parameters like diversity status, geography, detailed profiles and pricing.
It is true that searches must be performed quickly, but business decisions have to be made quickly as well – while still being fully informed. When a supply chain team starts their decision-making process with a qualified list of suppliers, they can seize the opportunity to access stock that might no longer be available if they delay too long.
Grappling with Globalization
Now that China has largely resolved the risk and impact of the coronavirus within their own border, they have once again become a global hotspot for PPE manufacturing. With the frenzy-fueled spike in demand, however, comes risks such as poor-quality product, falsified government approvals, and price gouging. Given the effort required to secure these scarce materials, having them arrive and not meet specifications or be of inferior quality is just as damaging as not being able to source them in the first place.
Suppliers must be trustworthy business partners, but supplier-related data must be trustworthy as well. As a result, it is not enough for supplier information to be up to date, it must also be validated and enriched by sources associated with sanctions, diversity certifications, and supplier reputation.
As we move forward through this pandemic, machine learning can help supply chain professionals keep their operation stocked with raw materials required for PPE as well as the products themselves. By using technology, organizations can find qualified suppliers fast and make informed decisions with confidence. They will no longer be left to select suppliers in a vacuum – instead, procurement professionals will be able to make smart supplier decisions and use analytics to generate faster and more valuable insights going forward. Knowing who you are buying from and what you are getting will help to solve today’s supplier discovery dilemma.