Predict, Not Just Respond

Apprentice Systems with Attribute Based Planning


Attributes are essential elements to define every aspect of a supply chain in as much granularity as needed. Attributes are found in every product, customer, supplier, equipment, tool, distribution center, method of transportation etc. To this end, attributes are dynamic constraints that can be defined by the users using Boolean expressions. An example of this would be: Only Supplier X is Qualified for Customer Y AND manufactured in NA OR Europe AND Processor Speed between 2 and 5 G. Having the ability to concatenate these constraints with “AND” and “OR” expressions allows the system to search for the right solution for every customer and order at the right time. Other examples of constraints or attributes are: shade of a fabric, wavelength of an LED, quality of a tool or equipment to be used for a certain customer of a certain status or region, materials used in a toy for certain countries.

As it can be seen, attributes can define the supply chain in great detail making digitalization a reality. In their absence, system lacks enough information to be truly a Digital Twin (per Gartner). Furthermore, by combining attributes and Boolean expressions with a search engine, that inherently looks for these user-defined constraints, an Expert System emerges capable of continuously learning from the users as an “apprentice” system. Users can add their own business rules using IF_THEN_ELSE expressions to teach the system describing what to do and when to do it. Obviously, this makes the system very malleable so that as the business changes the system molds itself to the new business rules and processes as needed without having to re-implement the system. For more detail on this topic and other innovations from Adexa click: ABP.



Leave a comment