There is an important question that is often requested, but not planned for when you are designing reports, dashboards, or other Information products.

It is a simple question to ask, and an easy one to answer, but data itself won’t support these questions unless they are designed in from the beginning-

“Where?”

At the risk of sounding like something that is too good to be true, by planning ahead on your data warehouse schema, and anticipating this question, you can get a lot of bang for the buck.

The “Where?” question could apply to everything from warehouse location for inventory, customer addresses, technician dispatching, critical infrastructure, or utility network connections.

To each of these examples, there are different kinds of “Where?”, ranging from X/Y/Z coordinates (for a warehouse) to an address (for shipping), to Longitude/Latitude (for critical infrastructure), or drive time (for technician dispatching).

The costs of acquiring spatial data are at an all-time low, and more or less free in many cases.  When you look around at the data sources, and the sensors embedded in so many devices now, the data is there, ready for you to collect.

Similarly, the costs of storing spatial data are pretty much free now.  You just need to spatially enable the storage tools, aka, the tables, views, and Data lakes/warehouses/silos.     In many cases, this is a configuration, not a customization, or at most a few more fields to a table.

Taking what we discussed last week, of beginning with the end in mind, it would help to anticipate how we might think about the “Where?” question.   We could be anticipating a map display (kind of obvious), or some sort of analysis, such as relationships or distances between any number of customers or suppliers.

Tying this all together, the trick is to collect data that is fit for purpose.

In order to create a point on a map, we need some sort of Geo-reference—the X/Y/Z coordinates.  This can be collected directly from something like a GPS, or it can be created by a (built in or standalone) Geo-Coding engine. Geo-coding is some sort of translation between where we think of something (a delivery address being a great example) to a coordinate (a Longitude/Latitude, displayed on screen).  In a building, these X/Y/Z coordinates would likely be related to a room, shelf or bin.

To understand the relationships between two or more items/people/locations, it may be necessary to collect more information, such as the network they connect to, including the rules of the network.   One relationship people immediately gravitate towards is Proximity— knowing the distance, direction or measurable time between two objects.  One long time reader of KJR (Hi Doug!) uses RFID based “finding” devices to look for routine medical supplies that might get lost.

Accuracy is also an important part of the fit for purpose discussion.  The right rule for most applications is “accurate enough to find it”.   For example, in many applications, it is important that the data can help you find the house on the right side of the street, but it may not matter if you have survey grade accuracy to the front door.   Don’t ask for precision that you don’t need.  Use only the minimum precision required that helps you meet the needs of your users.

Finally, the “Where?” question is an invitation for creative visualization.  Any application that returns some sort of text answer to this question is going to frustrate users.  While creating a map is one outcome, it doesn’t end there.  In Warehouse Management, the hot thing is “Pick to Light”, where a small device blinks, and guides staff to the exact bin or location containing the right inventory.   Simple database or warehouse enablement for spatial data answers a lot of answers, and helps you bring a lot more options to solve problems for colleagues.

Only now, this is easier than ever.