Fast Moving Consumer Goods Supply Chain Risk Management with Artificial Intelligence

OVERVIEW:

ITC Infotech provides technology solutions and services to enterprises all over the world, including banking, healthcare, manufacturing, and consumer goods. The latest advanced technologies are applied in their work.

Using anyLogistix supply chain software and applying AI (artificial intelligence), ITC Infotech built a supply chain digital twin for a fast moving consumer goods (FMCG) company that sells personal care products.

CHALLENGE:

Supply chains for consumer-packaged goods (CPG) are typically very complex, and the supply chain in question is no exception. It is comprised of factories, warehouses, hubs, and distribution centers all over India.

The performance of the supply chain is characterized by considerable variability, including:

Logistics
  • Demand variation
  • Transportation lead time variation
  • Production output variation
  • Supplier delays
  • Quality issues

Moreover, there is a long list of irregular disruptive events that can occur. Although it is possible to predict some of them, it can prove difficult to foretell their impact on the CPG supply chain over longer periods. Irregular disruptive events include:

  • Bad weather
  • Road closure
  • Factory/line closure
  • Unplanned holidays
  • Political crisis

Together, with incorrect decisions due to uncertainty, the factors listed above can significantly influence supply chain KPIs such as fill rate, excess stock, and transportation cost. ITC Infotech was commissioned to develop a tool that would help managers anticipate future supply chain behavior, get insights into the causes of potential problems, and help inform decision making.

ITC Infotech’s solution was to create a supply chain digital twin. To do so meant developing a simulation model that could get information about the real supply chain system’s current state and fully replicate its behavior.

SOLUTION:

The ITC Infotech specialists created the supply chain digital twin in anyLogistix, a special platform tailor-made for easier and more precise supply chain modeling. The model reflected the dynamics of the whole supply chain, including the secondary warehouse level.

Thanks to anyLogistix’s extensibility, the engineers used the multi-method simulation modeling capabilities of the buit-in AnyLogic simulation engine to create customized factory simulation components [the AnyLogic simulation engine comes as part of anyLogistix Studio Edition].

With the custom components, the engineers could embed the actual production plans and the Fair Share distribution algorithm to make the factory models behave like the client's real factories. The results could also then be fed back into anyLogistix. Unlike other simulation tools, anyLogistix also provided engineers with the possibility for pipeline consideration at the beginning of simulation.

Supply chain network and operation logic reflected in the digital twin

The inputs of the model included:

  • Data from ERP systems
  • Any available information about planned/unplanned events
  • Daily demand forecast made by the ITC Infotech Demand Sensing Tool (based on distribution and SKU combination data)

The output of the simulation was a 15-day detailed forecast of the most important CPG supply chain KPIs: stock level, fill rate, and logistics cost. After the forecast became available, AI was applied for the root cause analysis. This allowed users to understand the source of identified future underperformances.

AI supported automated root cause analysis for in case of low fill rate problem

The next phase of the digital twin functionality is now under development. It will provide users with a list of the best possible countermeasures. Such that, when a user chooses an action from a list, the simulation will send the necessary information back to the ERP for execution.

OUTCOME:

Using anyLogistix software, ITC Infotech engineers managed to develop a digital twin which now is able to:

  • Capture current snapshots of the client's FMCG supply chain business.
  • Predict the future for supply chain KPIs.
  • Navigate to the root cause of underperformance at specific points.

The long-term vision for the digital twin is for it to be an automatic tool that can suggest and execute necessary countermeasures to disruption. The project proved the efficiency of integrating simulation and AI techniques for the benefit of supply chain risk management.

Watch the video about this case study presented by ITC Infotech at the AnyLogic Indian Conference 2019