I’ve been talking and writing a lot about how structured content (for example, structured in DITA) can go far beyond technical publication departments and become a foundation for the intelligent content.

Considering the growing number of devices connected to the Internet and technologies that enable them to talk to each other and automate many processes, one of the areas where structured content can be especially beneficial is the Internet of Things.

Today, the number of devices connected to the Internet is 13 billion. By 2020, this number is going to increase to 20 billion. This means that product information will be also there.

In fact, the information itself will become a product. Just think about it: all these billions of devices connected with each other will be constantly gathering data, whether it’s data about the temperature, road traffic, your health, or availability of parking lots in the nearby mall. Some of this information will be exchanged just between devices, but some of them may be consumed by humans.

For example, imagine that a train got stuck on the railway and a service engineer needs to fix it. The train computer and sensors gather information about the issue and send it to the server. On the server, there is a content repository that contains all information about the train, including troubleshooting. This information is single sourced and conditionalized for various models of the train and various kinds of situations.

The information received from the train computer and sensors is analyzed and transformed to a publishing profile that determines what pieces of content should be retrieved from the content repository and how they should be assembled together. Once all required pieces of information are put together for this particular model of the train, issue, and environment, they are published and delivered to the service engineer (for example, to a tablet). Now the engineer can follow the instructions generated for this particular situation.


Moreover, this information will need to be presented in different shapes depending on the context. Maybe, if the troubleshooting process is long and complicated, it should be dynamically presented as a wizard. Or maybe the textual descriptions should be automatically converted to a process diagram that will be dynamically generated.

Or maybe, this information should be also personalized for this particular engineer based on the experience and skills. For example, if this is an engineer with less experience, then the troubleshooting procedure should include all steps plus some conceptual information. Alternatively, for an experienced engineer, only major steps of the procedure should be displayed.

A good news is that both the part that involves communicating between the train and the server and analyzing information received from the sensors and the part about context-based automated content assembly is real. Back in 90s, when I was working in industrial automation, we were gathering and analyzing data from gas pipelines to help operators make decisions. In our days, when the technology evolved dramatically, such a scenario doesn’t sound like a science fiction.

On the content assembly part, we have already done this. We’ve built a platform that allows third-party developers quickly build content automation applications. For example, one of the applications we’ve created automatically assembles an entire documentation set based on the bill of materials. Another application generates a flow chart from troubleshooting topics to visualize the troubleshooting procedure. One more application we are working on right now lets you generate a programming code in VHDL (the language widely used in the semiconductors industry) from technical specs written by engineers.

If you haven’t seen this video before, watch it now to get a rough idea of what can be done:


So content automation is already here. It’s just a matter of time when we’ll see it widely adopted. By the way, I’m going to talk about the potential that structured content opens for automation at the upcoming Information Energy conference which will take place in Utrecht, The Netherlands, on June 8-9. So if you are there, consider dropping by.

Leave a Reply

Your email address will not be published. Required fields are marked *