Digital Transformation – Control Towers and AI

In 2000-2003 (Around that time), MIT, MRA, IBM and some others were writing about advanced supply chain management platforms referred to as Control Towers (CT). As an EDI salesperson, I soaked that content up like a sponge. At that time it was a simple metaphor to help companies understand what they should work towards in terms of supply chain data. Fundamental digital transformation. Many companies have this capability now, whether coalesced into, or acquired as a single platform.

It is helpful when selling something as stodgy as pure-play EDI to have a relatable picture, or vision to share. A thing to work towards. Visibility of supply chain data at that time was a much larger problem. Even if only showing them the portion they are currently interested in, it helped. Sometimes it was integration with an ERP component, 3PL, or data warehouse, but the big picture has never changed: visibility provides an opportunity to measure and take action.

The most notable upside delivered in the last few years is without doubt Artificial Intelligence (AI). Initially CT was to deliver end-to-end visibility based on a “single version of truth”. Over time that mandate has morphed to include alerts, decision support, and most notably over the last few years, thanks to artificial intelligence (AI): the ability to manage many operations without human intervention.

What if any gaps are missing from your business

  • Is there a single version of the truth? Control Towers need to present data that is universally accepted as accurate and up-to-date. Without this your team is making decisions based on outdated, or just wrong information.
  • Is visibility end-to-end? Another way to look at this is determining if you have access to incremental information for each endpoint that touches any given product or service, including manufacturing, transportation, inventory and distribution at the item level.
  • Does your team receive alerts based on prescriptive logic? For example, if a supplier were to report an order will be late, do the various teams, their apps and exception reports get updated? The next point takes it a step further…
  • Can you optimize autonomously? For example, if supplier “A” reports via an EDI 855, that an order will be shipped late, can you automatically cancel that order and reroute product from a remote warehouse, or another supplier? More commonly, if a carrier misses a delivery appointment, can the new appointment be scheduled without a human’s involvement?

Supply Chain Control Towers make information visible, and accessible. They are there to help people take action on changes, or unplanned variances from different parts of the enterprise. As processes such as order-to-cash become established and automated, developing the controls to incorporate them into a cohesive control tower strategy makes sense.

Many companies achieve the same degree of automation using a hybrid approach using iPaaS and or ESB architectures and connecting to other networks – B2B (VANs), Unified Communications, and others. In between and connected throughout will be applications from several providers. The benefit there is the ability to phase in immediate value and minimize risk. Companies can fine tune with the help of multiple partners for far less upfront.

Going Forward

“Most agile supply chain wins.” This phrase becomes more true as companies develop autonomous supply chains. Your company needs to view the supply chain as multi-enterprise. Whether your company uses EDI, or some other B2B formats/methods is immaterial – that is just mechanics. Your company will use something, and for the next decade at least EDI will probably be in there somewhere.

What matters is that whatever the point-to-point integration is between enterprises, once the data arrives it doesn’t ‘sit’. The data should hit a data warehouse, or another platform so it can be re-used across teams, and applications. You want to see the value multiplied across the ecosystem more like a glob of protein. Nutrients for apps and teams rather than a static document. In the next few years, many leading companies will have transformed to the point that part of their supply chain is ‘usefully’ cognitive. Something to think about when planning for your next S&OP checkpoint meeting.