Several providers of transportation management systems (TMS) have invested in the user experience (UX). These suppliers have worked to improve both the ease of use, the intuitiveness of solution, and the ability of power users to do their work efficiently, for example, without having to have multiple screens open. In TMS, that means a focus in particular on the role of the dispatcher. TMS solutions have added a Google Maps look and feel to the dispatcher landing page, and most TMS solutions have drag and drop to allow dispatchers to drag loads from one truck to another, or a driver from one load to another, and then have the plan automatically update itself.
SAP has also invested in providing their planners with Gantt Charts to view the scheduled loads. Bill King, the Director of Supply Chain Execution Solution Management at SAP, made the point that with a “Gantt Chart dispatchers are not forced to trust the output of the optimization engine. It appears to us that many transportation planners must be from Missouri, the ‘show me’ state, because you have to show them before they believe it.”
The idea that a supply chain solution should not be a “black box” solution – a solution that forces planners to blindly trust the output of the optimization engine – has been around for close to twenty years. In transportation, the need for this has been glaringly obvious because of all the parameters that can go into a “solve.” The goal of the solve is to hit service commitments at the lowest cost. But historically the output of the engine might not be totally feasible for a whole host of reasons. So for example, Gina does not want to be on the road for more than seven days in a row, or George does not have enough Hours of Service (HOS) left to do a particular delivery and arrive on time.
In addition to driver constraints, there can be constraints surrounding equipment (two trucks are in the shop for maintenance), customers (an important customer wants the deliveries to be handled by Dave), products (Hazmat products can’t ship with produce), and sites (you can’t deliver to this location after 6 pm). Unless the optimization model fully understands these constraints, and historically they have not, the output will often not be fully feasible.
This is why TMS solutions need a planner’s workbench. These workbenches give planners the ability to override an engine’s output and manually manipulate loads prior to releasing a truck and a driver to a route.
But while recognizing that not all plans produced by an optimization engine will be fully feasible, and that there will always be a need for planners to adjust the output, leading TMS suppliers are making a real effort to have solutions capable of modeling virtually all constraints.
Derek Gittoes, the Vice President of Supply Chain Management Product Strategy at Oracle, said “the goal really should be to try and get away from replanning. You should not get to the stage where you are assigning loads and find our you can’t cover all 20 shipments.”
The Oracle solution is a robust solution, but the goal of being able to fully model all constraints is never ending. “In our fourth quarter release we are expanding our driver management capabilities.” This means modeling when drivers are available, when they want to work or not work, their qualifications – whether they are qualified for Hazmat loads, for example, and what type of driver they are (over the road or a local driver, for example).
Fab Brasca, the Vice President of Solution Strategy for Intelligent Fulfillment at JDA, pointed to the difference in routing a private fleet versus loads that are tendered out to common carriers. “In setting up a fleet it is important not to treat it like a commercial carrier, the fleet needs its own behavior.” Things like asset types – a reefer versus a flatbed, for example, are the same from a scheduling perspective in either type of solution. But in other areas, the fleet model has to be deeper. So for example, if you tender out to a carrier, the carrier is expected to make sure that the driver has enough Hours of Service to make the trip. For your own fleet, you need to know what the HOS situation is for specific drivers in your fleet. You also need to model the home domiciles for drivers and things like the locations where trailer swaps can occur. A trailer swap occurs when drivers meet in the middle of long routes heading in opposite directions and switch trailers.
But Mr. Brasca also strongly made the point that for shippers with private fleets, fleet routing is not the be all and end all. “Transportation optimization requires a holistic view across all modes. JDA’s approach is that as opposed to building specific fleet capability, we come from a paradigm where you look to plan across all modes concurrently.” Further, when it comes to road transport, the model should be capable of planning fleets in conjunction with common carriers. And for multimodal shipments, the solution should be capable of making “very complex comparisons through multiple legs.”
One of my jobs as an industry analyst is to put together supplier selection guides that contain configurable criteria that companies can use to select a solution that is right for their business. This year’s TMS supplier selection guide was enriched with much more fleet functionality than the previous version.
In putting together this guide I came to the realization that if you understand a solution’s model, what constraints and parameters the solution models, and how granular that model is, you also understand the functionality that solution will be able to deliver. So in a request for proposal (RFP), for example, you might ask for functionality such as support for road trains – a cab pulling multiple trailers. Or you can ask how the solution models equipment. For example, does it model cabs and trailers separately? And what kinds of regulatory details are included in the routing map?
In the title I asked the question, “What Fleet Routing Solution is Right for My Company?” Clearly the user experience and the depth of the supply chain model are two key things to look for.
