Location and the Internet Of Things

Preface

A lot of discussion is going on about the internet of things, and e.g. in Finland and Germany the Industrial Internet is creating interest as the most likely sub-set to see the first implementations. In a way, there have already been implementations by some of the innovative companies years before the phenomena was categorized and called "Industrial Internet".

The simplified concept of IoT and the Industrial Internet is easy enough to understand; by enabling everybody and everything with connectivity and automating processes our lifes become easier and business processes more efficient - with improved sustainability.

Where's the cake?

So, what is actually the driving force behind all this that is good and beautiful? In my mind, the real benefits come from giving the connected devices and humans better information to do better decisions. So how do you make better decisions? The data that drives your decisions must be easily understandable, valid and timely – but also relevant. No data set is going to be useful, if processing the data set contains too many irrelevant inputs and attributes and in short is too complex to be quickly processed.

Of course, the restrictions of complexity are mostly relevant to us feeble humans – thanks to the connectivity the devices (and humans) of IoT / Industrial Internet can tap into a wealth of processing power. What the machines lack is the ability to understand relevancy an relation – they must be programmed to understand the physical environment, which in a way is the essence of GIS.

Let me make a few examples.

Let’s assume that an industrial equipment, e.g. a crane or an engine has been analyzing it’s condition and decided to order a predictive maintenance along with some spare parts. For the whole process to be truly automated, the machine-created service request would then need to be assigned to the best available service person, from the closest (or cheapest) possible location with the correct parts to the right place at the right time. The only human touch could be that of the service person accepting the call with e.g. his mobile device. For this kind of automation to happen, the equipment (or the logic controlling this) needs to know the actual locations of persons, their availability, their competency and their distance and transport cost to the service site. The same things apply to the spare parts as well. How would you do all this without analyzing location?

For machines to be able to think and optimize better than humans, they need to understand the geography as well or better than humans.

Think of an analyst trying to understand the patterns of machine malfunction. Due to the huge amounts of (sensor)data she needs to be able to group the relevant data. One way of grouping all this data is to do spatial analytics – i.e. study patterns by geography. Where are most of the malfunctions happening? Are they grouped to a specific area? What attributes are different in that area compared to other areas? Is there a way to explain the root cause by examining the business in that specific area? In sales, wouldn't you be glad to analyze which services and brands sell in which locations?

Location can be used as a common nominator when working with complex, multi-source, unrelated datasets with structured and unstructured data -enabling new insight to existing information.

The cake is not a lie

We all start learning geography from the day we a born. The areas around electrical sockets are verboten for my 9 month old daughter. Maybe your spouse has suggested that some of the areas containing pubs and bars are not good for you during the mother-in-law visits. When a supply chain manager is considering a new supplier from Ukraine, she immediately, intuitively knows some of the strengths, weaknesses, opportunities and threats that relate to that country. Maybe the geopolitical situation will make this supplier vulnerable? Maybe they are willing to sell cheap because of that? Maybe we can grow our foothold in the area?

This is the type of information that us humans have gathered during our lives and use every day without putting much too effort to do so. As the amount of data and the complexity of decisions become more vast we will sooner, rather than later later, want to use both analytical tools (location analytics) as well as visual aids (maps). Some information about an area is stored in to our brain, but do we have the relevant information? Are we up-to date about the demography, transportation network and other information in that specific area - or should we assess the situation with the help of maps containing information relevant to that decision?

Maps are merely an interpretation of the data, and when considering your location strategy in IoT and Industrial Internet it is critical to understand that maps are for people, location information for machines. And in many cases, people don't need the maps either, they just want the result of an analytic process (In a 2 hour drive, how many possible customers are there from where I am now?).

So, how is this relevant to your business? Efficient use of maps and location analytics provide a number of ways to save money, increase your performance and optimize your workflows – but only if you are willing to make the first step and start exploring the possibilities. Once you do, you can move on from exploring to using, creating and sharing relevant maps throughout your organization and business processes. (Some real-life examples of the benefits brought by use of location include Bank Of America and General Motors).

Conclusion

The amount of business-, sensor-, market-, customer-, you-name-it –data is not going to do anything else but grow and then grow some more exponentially, so the time is now to make sure that you are not missing out in one of the biggest assets in that data – location. If you have already implemented location into your business, congratulations! Now you need to make sure that it's being used efficiently throughout your organization, not just a single department or a business unit.

TL;DR summary

• Machines don’t have the same analytical capabilities that humans do
  • Machines need to be told about the physical environment using e.g. location information
• Humans are poor at understanding large amounts of seemingly irrelevant data
  • Location can be used as the common nominator between different data sets
• If you don't invest in location, you may be missing out on some substantial benefits.