From SCDigest's On-Target E-Magazine
June 1, 2011
Logistics News: PepsiCo Brings Automated Wave Management Concept to Transportation Planning
Existing Process with Huge Volumes wasn't Sustainable, PepsiCo's Ukidwe says; New Optimization Cockpit Automates Order Selection and Optimization Process Across 10 Divisions
SCDigest Editorial Staff
PepsiCo, one of the largest shippers in the US across its many division, has taken an innovative approach to transportation planning and optimization that allows it to automate several existing steps and decrease its freight spend in the process.
Pepsico, which in addition to its traditional soda group runs such businesses as Tropicana juices, Gatorade, Frito-Lay, and other divisions, also recently purchased its largest bottler, adding further to its network and transportation complexity.
SCDigest Says: |
 |
| Because of the more comprehensive order selection process based on rules and the auto capability to puts low quality orders back into a future optimization run, PepsiCo gets better optimization results. |
|
What Do You Say?
|
|
|
|
All told, PepsiCo manages at least $2.2 billion in freight spend, and some 1.5 million shipments per year, including private fleets, dedicated carriage, direct store delivery, refrigerated trucks and other requirements in addition to straight truck load shipments across its huge supply chain network.
According to Animesh Ukidwe, a transportation solutions manager at PepsiCo, as the complexity of the network grew through acquisitions and more areas of the company being brought into PepsiCo's load control center (LCC), the number of rules, requirements and constraints by product, channel and customer proliferated along with the basic challenges of the immense volumes being managed.
As its load control center evolved over the last 10 years from managing a single division to 10, with corresponding increases in volumes and complexity, the process of optimizing available loads and moving them through execution became overwhelming. Like most LCCs, PepsiCo planners with expertise in a given region or division would select orders from an available pool, run them through the transportation management system for optimization, and then send them on for "local execution."
This somewhat manual planning process, which depended heavily on the tribal knowledge of individual planners, was "not sustainable," Ukidwe said recently during a presentation at the JDA Software users conference. PepsiCo has been a long time user of the i2 Technologies' TMS solution; JDA acquired i2 in 2010, and is in the process of developing a new release that will combine the features of the i2 and existing JDA TMS products.
The existing process was error prone, not scalable other than throwing more people and dollars at the challenge (not a very popular strategy today), took too long to execute, and often involved spreadsheets and other "unsupported apps" used by planners on the side, according to Ukidwe.
New Optimization Cockpit
Ukidwe led the effort to go to a much more automated solution, building what he calls the Optimization Cockpit.
The new system, which is basically built on top of the existing TMS, works much like advanced wave management approaches in a distribution center. It lets planners build an infinite number of rules for selecting orders from the general order pool across the company, apply various other requirements and constraints, and then have the system either kick-off that run automatically or have a planner do it with the click of a button.
This is much like a wave planner in a DC builds preset order selection rules to grab available orders to drop to the distribution center floor. Ukidwe said that in practice, about 20 different variables are used to select the different order pools for optimization.
(Transportation Management Article Continued Below)
|
Those optimization runs then usually occur automatically, based on the timing defined within the Cockpit for that particular order set. For example, the Frito-Lay inbound raw materials plan is run daily; others may be run more or less frequently. Now, instead of manually selecting the orders to be optimized, the first time a transportation planner at PepsiCo usually sees what the system-generated plan looks like is after the optimization has already completed. (A few difficult order sets are not included in the Cockpit runs due to their planning complexity.)
