Optimizing shipping logistics in a time of change

Within logistics, shipping is a vast and delicate ecosystem. Over the last couple of years many people were directly impacted by complete production shutdowns, huge and unexpected swings in consumer demand, lack of labor at ports, a shortage of shipping containers… just to name a few!

Addressing challenges with business analytics

To help with some of these challenges, my company Spitfire Analytics has been working with a global retail organization that is responsible for one of the largest independent shipping networks in the world. This client already used IBM Planning Analytics with Watson to provide inputs to a demand plan and combine it with various input factors – allocation of volume from a business unit to a specific address, filling rates of containers, etc. However, most of the calculations were done in a traditional relational database with week-level data granularity that took approximately 4 hours to run allocations.

When market disruptions occurred, their demand plan became so much more vital to answering critical questions, such as: “Do we have enough carrier capacity to transport our forecast volume? Do we have enough handling employee hours available at a warehouse to unload the inbound stock? Can we give our customers what they want?!”

The decision was made to move to a more continuous planning cycle as the need to align the traditional mid-term plans (monthly granularity) and compare it the operational plans (weekly granularity) became vital. The only option was to start breaking the data down to specific days and dates. The existing relational database logic was already taking 4 hours to perform 1/7th of the calculations that would be required – so that was a no-go.

Leaning into the benefits of IBM Planning Analytics

This is when we decided to do everything in IBM Planning Analytics, which enabled:

  • A single platform for user access
  • ETL imports from multiple source systems
  • An advanced UI to easily process and calculate vast allocations and export all that data to be picked up by external applications

All these benefits resulted in getting the total processing time to under 1 hour. Now users could see results of their plans quick enough to make decisions and act efficiently.

Through the scalable nature of IBM Planning Analytics, we were able to handle algorithm sparsity exceptionally well. When routes introduced multiple stopping points, carriers were allocated volumes, and lead-times were applied to provide a receiving date, resulting in approximately 300 million records being exported to a Database table every night.

Now working with that same retail customer, we are at a point of producing multiple other applications that take those outputs and assess and project where they may encounter a bottleneck. For the long-term tactical planning process, this customer can start making fast, data-driven decisions like whether to temporarily rent external warehouse space, build a completely new warehouse, enter into new shipping contracts, or make new staffing changes.

Lessons in optimizing logistics

Ultimately, we found the business needed to change its way of working due to external factors outside their control. We’ve been hearing this consistently from many customers as a result of unexpected changes and market shifts that can impede the growth of a business. Accordingly, we’ve found a good way of working that involves identifying the root cause of problems in the planning process and quickly focusing on developing and implementing a planning solution that works alongside representatives from the customer’s organization – often in as little as 5-10 days of consultancy time. This leaves the customer not only with a better planning solution but with the skills to start expanding that solution and use case for themselves to handle the next set of business priorities.

By working with a strong planning analytics solution, organizations no longer need to require someone to be versed in multiple coding languages and various skills like Macros in Excel. Now, teams across any department – especially financial and operational teams – can work from a single source of truth to streamline planning, reporting, and analysis to manage performance and build alignment across the enterprise.

Ready to learn more?

The post Optimizing shipping logistics in a time of change appeared first on Journey to AI Blog.