How to build a GIS-based Decision Support System for logistics using OpenStreetMap
Making decisions, whether strategic or operational, requires processing of large data arrays and transforming them into actionable insights. In the current era of Big Data, humans are unable to handle even a fraction of available information. That’s where a GIS-based Decision Support System (DSS) comes into play.
What is DSS?
To put it simply, DSS is a system that supports business decision-making processes using fully computer-powered or hybrid approaches. The latter combines machine-generated results with our personal interpretations and consequential conclusions.
A properly designed decision support system is an interactive software that aggregates available information and presents it to the end-user in various formats: sales figures comparisons, revenue forecasts based on archive data, interactive geodata visualizations etc.
The number of potential use cases is rather unlimited, considering the constantly growing digitalization of each and every business sector.
What is spatial DSS (SDSS)?
Spatial decision support system extends DSS capabilities by combining a DSS and a GIS (geographic information system) into a single, integrated software that helps predict possible outcomes of business decisions that depend on geographical data.
My aim in this article is to focus on building a GIS-based decision support system for the logistics sector, so we’ll be mostly talking about SDSS. Let’s begin.
Systems for decision support in logistics
Transportation industry relies heavily on processing spatial data in order to deliver valuable insights to the decision makers. As I mentioned before, sometimes the analyses are fully-automated and use backend algorithms to calculate or predict:
- transportation costs to and from selected destinations;
- fuel expenses based on route’s elevation (driving shorter uphill path might consume more fuel than driving longer route with smaller elevation or lack thereof);
- optimized routes for delivery (for example, UPS tries not to turn left when completing deliveries to save fuel, drive shorter distances, avoid crashes and decrease CO2 emissions);
- and more.
At the same time, hybrid decision support systems can also be of use when it comes to running proper logistics operations:
- Departures and arrivals times suggestion based on predicted traffic intensity or weather conditions, road limitations for trucks (weight limits on bridges, height limits in tunnels etc), or cost comparisons for various transportation methods may provide freight forwarders with better planning capabilities.
- Product demand visualization can help chain managers to properly oversee their deliveries, especially in situations when various crises hit and call for rapid yet concrete decisions to be made on the spot.
- Lists of available truck parkings along the selected route can help truck drivers plan their journey in a more efficient way, taking their minds off the need to manually search for places to stop and rest.
The above are just a few examples of decision support systems that can be developed and implemented within your organization. Now, what happens if you are actually thinking about building one?
Building a decision support system using OpenStreetMap
Before we dive deeper into the process of building your own spatial DSS, it is worth making a short introduction to OpenStreetMap, if you’re yet unfamiliar with it.
OpenStreetMap (OSM) is the world’s largest open-source editable geographic database. Essentially, it provides you with access to fairly accurate spatial data for free, which you can then utilize as a foundation for building your custom DSS.
If you’d like to learn more about OSM itself, you can check my latest map APIs comparison to find out how it compares to Google Maps Platform and Mapbox, or read Magdalena’s post on how OSM helped one of our clients to save approximately $420 000 on monthly API usage costs.
When approaching the topic of building your own SDSS, it’s best to start as if you would with any project – product workshops. It’ll allow you to lay out all the requirements and see whether your initial technical assumptions remain valid.
Once all is set and done, we can move to the most complicated phase – development. If you’re working on a limited budget, we advise starting with a data visualization system.
In general, spatial data visualization systems are cheaper to develop and implement than a sophisticated, fully-automated computer system. They can also serve as a PoC that later transforms into an algorithm that doesn’t require human interference. It took 10 years for UPS to get it right with ORION – their routing software that favours right-hand turns. Speaking of which…
Why does UPS avoid turning left?
UPS – a worldwide shipping, receiving and supply chain management company, unveiled why their delivery drivers try to steer away from turning left. It’s rather fascinating, if you ask me.
First of all, it’s important to mention that this solution is based on right-hand driving system, and can be implemented in left-hand driving countries like the UK or Japan. You would simply have to avoid the opposite – turning right.
But why does it matter how delivery drivers turn, though?
Well, let’s start with the US National Highway Traffic Safety Association report that states that 61% of crashes that occur while turning or crossing the intersection involve left turns. Right turns are responsible for a measly 3.1%. So, for starters, by avoiding left turns, we avoid potential crashes.
Next, ‘a left-hand turn is also less fuel-efficient, because your car's idling [is] longer, which is also not good for your vehicle.’
— Jack Levis, Senior Director of Process Management at UPS
And, while avoiding a specific type of turn might sound confusing as hell, the numbers speaks for themselves:
- UPS makes 18 000 000 deliveries daily;
- ORION analyses 250 000 000 address points daily;
- ORION performs 30 000 route optimizations per minute;
- 6 to 8 fewer miles driven per each route;
- ORION saves 10 000 000 gallons of fuel per year;
- This saves from $300 000 000 to $400 000 000 annually in fuel, vehicle running costs and salaries;
- ORION saves more than 100 000 000 miles in routes a year;
- This leads to a reduction of 100 000 metric tons of CO2 emissions per year;
- Which is an equivalent of taking 21 000 delivery cars off the road.
Technology behind UPS ORION
Just to be clear, UPS doesn’t simply ban turning left. Driving in circles by always turning right wouldn’t be the most effective way of delivering goods. I think.
Instead, UPS built a system that analyses an entirety of a selected driver’s route, identifies all the left turns and eliminates the ones that can be avoided.
To manage logistics of delivering a high number of orders, UPS needed an effective tool. Google Maps, for instance, has no concept of not making some type of turn, it simply shows the shortest or fastest route to the destination, which in turn (see what I did there?) wouldn’t satisfy UPS’s needs.
This led UPS to creating their own custom maps of the locations where they operate. Their system knows about parking lots, road limitations (like the one I mentioned above with the truck weight limits), private driveways etc. All of this geospatial information provides ORION with the data required to optimize logistics routes. And while it is undisclosed whether UPS utilizes OpenStreetMap, it is safe to say it can deal with the task at hand.
Trans.eu and OpenStreetMap
Trans.eu is one of the leading logistics platforms in Europe and Asia as well as our long-term partner. With more than 100 developers from our team involved in developing various parts of their system, we’ve also helped them to build a hybrid decision support system, amongst other things.
Based on OpenStreetMap data, we’ve created a number of spatial geodata layers that help Trans.eu end-users plan better logistics routes and transport goods all over the continent.
Custom geographic data, in the likes of aforementioned road limitations, can be drawn on maps in real-time, offering freight forwarders operational insights and strategic advantages over their competitors.
We also use OSM data to power a machine-learning model that calculates transportation price recommendations based on route data, historical prices, available offers on the platform etc.
Being able to utilize open-source geographic data and combine it with other data sources allows you to build complex logistics decision support systems. Eventually creating your own ORION and saving hundreds of millions of dollars.
As you can see, building a location-based DSS can bring numerous benefits and provide your organization with data-enhanced decision-making processes.
Utilizing open-source tools, in the likes of OpenStreetMap, can help you save large sums of money and create a unique system to outperform your competition.
If you’re not sure how to tackle the process of developing your own decision support system, drop me a line at email@example.com, and I’ll connect you with our geospatial experts.
That’s it for today. Ciao!