In this post, we’re going to lay out best practices for data management in logistics businesses of all types - aviation, ground transportation, or anything in between - and whether you’re doing crew scheduling, vehicle routing, maintenance scheduling, or any other planning related to logistics.
Note that we use the term “schedule” to mean any activities that are associated with a time and/or date. It is meant to be generic to all planned activities.
The Minimum Data to Track
If your goal is to use the data in future to make informed business decisions, improve efficiency, and potentially look at implementing some optimization, at minimum you should be tracking time-stamped schedules.
This will at least show the progression through the planning process. The data should ideally be in some sort of flexible format (spreadsheet is better than PDF; .csv, .json and other similar formats are best).
Operated vs. Planned
We talked previously about the issues when this isn’t tracked, so the next step is to add the actual operated schedule to the iterations of the planned schedules.
Tracking what was operated vs. what was planned is key for identifying schedule changes that happen repeatedly, which is where a lot of efficiency improvements will be found.
In addition, it will give insights into how much time and money is spent each year on disruption management, last minute changes, understaffing/overtime costs, and more.
Standardized, Flexible Format
We’ve mentioned the issues that come up when data is stored in a particular software without the ability to import/export or move around.
Ideally, all data you track should be in a flexible format that isn’t tied to a particular software. Not only does this protect against data loss when software becomes obsolete, but it’s much easier to manipulate this data later to be useful.
Keeping data agnostic from any given software tool or format ensures that it will be available and easy to analyze later.
Context is not only difficult to remember in the future, but we tend to adjust our own reasoning when looking in hindsight.
Notes about why particular changes were made to a schedule, disruptions, and general comments will give context to a schedule or dataset. This makes it much easier to evaluate and use in decision-making in the future.
Even if you have a full dataset, without the underlying reasons for changes, it becomes difficult to detect patterns and trends that could be a source of efficiency improvement.
When making future scheduling decisions, a canceled flight every Friday for the past month that is coincidental is much different than one client who always cancels. Yet this will appear the same in the data, if no context is given.
In your business, ideally you should track all of the following:
Time-stamped schedules through the planning process,
Final operated schedules,
In a standardized, centralized, software-agnostic format like .csv or .json,
And add notations to the schedules when changes are made with the background and reasoning.
If you manage to do all these things, you’ll be ready to make informed, data-driven business decisions that improve the efficiency, and ultimately, profitability, of your business.
You’ll also be ready to implement optimization technology, and make transition to new tools and software much easier.