By accepting you will be accessing a service provided by a third-party external to https://buyersmeetingpoint.com/
Webinar Notes: The Dawn of Risk Management: Harnessing the Power of Watson
These notes are from a June 23, 2015 webinar hosted by Sourcing Industry Group and presented by Louis Ferretti, a Project Executive at IBM. While only SIG members can view the recording on demand, you can catch Ferretti at their Global Executive Summit in October.
I knew I wanted to attend this event as soon as I heard Watson, the artificial intelligence computer that competed against two of the best ever Jeopardy! contestants in 2011 and won, would be featured. If that kind of AI could be applied to supply chain risk management, just think of what might be possible! In this case, IBM presented from the buy side perspective, although many companies are familiar with them on the sell side. Watson was applied in the management of IBM’s own spend.
One of the key values to applying Watson to supply chain risk is the fact that it can handle unstructured data (meaning free text in articles and social media) rather than being limited to data in tables. The ultimate goal is to uncover correlations that can be used predictively to improve decision making. In addition to unstructured data, Watson can handle such large volumes of data that it is possible to identify secondary correlations that are not apparent to human analysis.
IBM’s Enterprise Risk Management division looks at pervasive risk at the corporate level rather than limiting its view to supply chain. That being said, procurement/suppliers was rated by IBM as one of their top 20 risk areas. An interesting side note is the fact that getting visibility still outranks supply chain risk as the top procurement concern, demonstrating that while understanding of risk precludes mitigation planning, visibility into the supply chain precludes understanding of risk.
The objectives of what IBM calls the Watson ‘Risk Rover’ project are deceptively simple. Watson starts by taking unstructured data and plotting an expected path. Then the questions are:
1. Do actual events track to that path?
2. If yes to #1, does the path of the disruption affect the supply chain?
Over time, IBM can then look into the probability that a series of indicators will result in a disruptive outcome. This is critical because while disruption comes with costs, so does risk mitigation. Taking steps ‘just in case’ is too costly to be effective.
Despite the seeming genius of Watson’s supply chain risk recommendations, apparently IBM’s procurement team was hesitant to rely upon it. That is, of course, until March 11, 2011: the day that the tsunami hit Japan and seriously disrupted the global technology supply chain.
The Risk Rover project will conclude at the end of August, after which time we will hopefully get to hear more from Watson and the IBM supply chain team.