8-steps System Integration Model

8-steps System Integration Model

BEMS, BMS, BAS, EMCS, and this list goes on and on. Welcome the acronym-filled wasteland know as building automation. What you are about to read may seem high-level, but I promise if you follow these steps you will take your skills to the next level. Let’s begin.

One of the persistent challenges I hear from my audience is around the topic of systems integration. Automated Buildings readers you’re in luck! Today I am going to give you a high-level overview of my 8-step process for systems integration.

The Systems Integration Model

Using this process I have personally done some of the most complex integrations in the smart building space.

OPC integration bringing 13 different Building Automation systems AND THEIR DATABASES into a single front end. Check!

Tying together Lighting, Physical Security, Video Management, BAS, Maintenance Management Software, and Google Calendars for scheduling. Check!

Writing applications that consume XML data feeds from clinical systems and then convert this data into a BAS system. You Bet!

What you are about to read may seem high-level, but I promise if you follow these steps you will take your skills to the next level. Let’s begin.

Step #1: Define the Business Challenge 

This may seem like an odd step especially for technical folks. The reality is without defining the business challenge you will never get adoption and support from all of the different stakeholders.

Step #2: Create the Use Case 

Now that you have defined the business challenge you need to create the use case. I prefer to use the UML modeling method for my use cases but you can use whatever method you and your customer are comfortable with. The key point is to:

·         Capture what the outcome is

·         How the outcome is reached

·         Who is taking the action

·         What action they are taking

Step #3: Identify the Systems 

You now have a functional use case. We will now begin to dive into the technical aspects of integration. It is here that you will go and dig into the use case and identify the systems that are being used. You need to be very detailed in this step. Often times there will be systems that are being used that are not called out in the use case. You want to:

·         Identify the systems in the use case

·         Identify any systems required but not detailed in the use case

·         Identify any people in the use case

Step #4: Detail out the Data Flows

Now that we have the systems detailed out we need to define our data flows. Where is data flowing? I like to use Crow’s Foot notation to show how my data flows are laid out. Essentially what you are trying to do at this point is to detail out:

·         Who is the master system and who is the slave system

·         Which way data flows in the integration

·         If the data will be one-to-one, one-to-many or many-to-many

Step #5: Build the Data Model

Great! We know which way data is flowing. Now we need to detail out our data model. 

What points do we need from each system? 

How will the points be formatted? 

What protocol will the points use? 

It is here that we detail out our “data model”. For this I like to use the UML class diagram. It may seem like overkill but a class diagram is a great tool to avoid having systems that won’t map to one another.

In the class diagram I map out:

·         The points that will be available at each system

·         How the points are formatted

·         The frequency of the points being sent

Step #6: Map out the Network Connections

Naturally we need to send data to and from our systems. The way we do this is via the network. Now you may be wondering why I waited to do a network map until now.

The reason I waited is because I wanted to know which systems needed to talk with one another. By first detailing out my data models and my data flows I defined who really has to talk to who. 

Many folks will look at the use case and immediately start mapping out their network connections. This results in people missing key systems that the use case did not actively mention.

Step #7: Prepare the Physical Integrations

It is at this point that we will want to begin to prepare the physical integrations. Here we will setup any integration cards, protocol gateways, etc. At this point in the process you should have a clear path to finishing your integration. This step usually includes:

·         Setting up any integration gateways

·         Configuring IP addresses

·         Working with IT to get any routing setup

Step #8: Implement the Integration

This almost isn’t a step! You simply need to implement your network map and data model. Really at this point you simply need to show up and coordinate the other vendors. It’s funny to me because this is the step so many people try to do first and then they wonder why this step is so hard. 

Honestly, if you’ve done each of the steps up to this point this step should be a non-event.

Artical Publish by Safe Secure Magazine in May 2019 edition.