Meta-Morphing Mashup for Self-Creation

I am sure my readers think I simply make up these column titles as I attempted to describe the various pieces of our ongoing digital transformation.

All these words come from articles, interviews, news releases that appear in our monthly global online resource for large building automation which has evolved to become a staging website for industry thought and evolution. The words meta-morphing are no exception, but you will have to read the full column to find their source – and to understand why they make my head explode!

My last column, Transparency and Digital Twin is the New Deal, seems to have captured the imagination of many. This column continues to explore how we might achieve those qualities and the transition to artificially intelligent, self-learning systems, and a world of self-creation.

The Mashup started with a significant announcement earlier this month, ASHRAE's BACnet Committee, Project Haystack and Brick Schema Collaborating to Provide Unified Data Semantic Modeling Solution. From the release:

The ASHRAE BACnet committee (www.bacnet.org), Project Haystack (www.project-haystack.org) and Brick Schema (www.brickschema.org) announced they are actively collaborating to integrate Haystack tagging and Brick data modeling concepts into the new proposed ASHRAE Standard 223P for semantic tagging of building data.

ASHRAE Standard 223P: "Designation and Classification of Semantic Tags for Building Data" provides a dictionary of semantic tags for descriptive tagging of building data including building automation and control data along with associated systems.

By integrating Haystack tagging and Brick data modeling concepts with the upcoming ASHRAE Standard 223P, the result will enable interoperability of semantic information across the building industry, particularly in building automation.

What the heck does that mean, and how does it help get us to some of the next steps in our transformation?

In this article, BAS Industry Collaborates On Semantics, our contributing editor Therese Sullivan (who writes on many other industry topics), shares her thoughts on the New Deal and starts to unravel the significance of this announcement:

When competing open-standards are allowed to proliferate, you get market confusion. So it is great news that those working on semantic interoperability are now actively collaborating under the umbrella of the new proposed ASHRAE Standard 223P for semantic tagging of building data. This collaboration includes the ASHRAE BACnet committee, Project Haystack and Brick Schema.

Project Haystack was early to organize and tackle the challenge of semantic data modeling for buildings, but, awareness of the issue is now mainstream. Claire Rowland, Interaction Design Director, at the well-known global design firm Method, wrote this post about designing a UX for connected embedded hardware, including the following paragraph concerning semantic naming:

“.…At the most basic level, it’s important to give the same features the same name across all interfaces. But even this can be a challenge when the same app needs to support multiple versions of legacy hardware. Perhaps device designers had different ideas about whether ‘auto’ or ‘timer’ was the ‘best’ name for the heating schedule function.”

Rowland, like UX designers everywhere, wants what Haystack cofounder John Petze describes here:

“The reason that you can point a web browser at someone else’s website to see what they have to say is that we have all agreed upon how website data should be marked-up or tagged. You and I don’t have to have any pre-negotiations to make that happen. We need the same thing for data coming out of systems and devices. We should be able to get all the data, and it should be easy to interpret when we receive it. It should be marked up with tags that describe what it is. Project-Haystack has been working on that challenge since its founding in 2011.”

By calling the things in our world by a common name that is machine readable many other parts of the digital transformation puzzle start to fall into place. In this abstract, Alper Uzmezler has put out an open invitation to join him in developing Project Sandstar with the goal of making it easier to semantically tag points and to implement controls. Three key points to how all this will work:

• Sandstar GPIO engine to handle real/time connectivity to I/O for sensors and actuators activated based on haystack tags
• Sedona to provide real-time programmable control logic  https://sedona-alliance.org
• Haystack to semantically model and configure the control points https://project-haystack.org

From the abstract:

Methods:

Device template is created in order to abstract system calls to GPIO, I2C, UART and other hardware level communication. Within this file we have utilized a tag called “channel”. A zinc in the form of grid with records where “channel” tag which is type number is generated. Within the zinc file we are forming all tags necessary to convert raw data into cur val data form. This consists of tags that defines programming data type, port definition, sandstar engine, localization conversion and last regular haystack tags are defined for application abstraction. Sandstar C/C++ engine consumes both files and communicates to both Sedona and Haystack rest api.

Results:

Abstraction in both hardware level and point definition level enables us to generate portable Sedona code. With the abstraction of IO, sedona code can be written in a way it covers all potential input, output and setpoint types. With the power of haystack ops (Rest api), we can add a record to the sandstar engine thus enabling the branch of the Sedona code that is waiting to be activated. We call this feature meta-morphing programing. On the roadmap, having haystack client in sedona will help us to have P2P device communication along with historical data and analytics based control.

Conclusion:

Sandstar project will change how we think of DDC. Major improvements are hardware independent sedona code, historical data based control logic, driver abstraction via haystack can be achieved now. With the improvements to haystack ops where Sedona components can be created changed deleted and linked, artificial intelligence can be utilized to generate and improve upon human generated DDC code.

Therese -- wearing her Project Haystack Hat -- provides this key insight, that Alper is open sourcing the general-purpose I/O engine as well as tag definitions that he developed within his company, Anka Labs.

He is championing a Project Haystack working group, inviting other programmers in the Haystack community to improve and secure the architecture and engine and to expand on the initial tag definitions.

As Project Sandstar evolves it is sure to attract the interest of programmers and product developers from the larger IoT and “maker” communities. With their help, there is potential to expand Project Sandstar to include tagging dictionaries for sensors, output devices, and control applications, well beyond HVAC control.

Alper presented an overview and a demonstration of Project Sandstar on a recent ControlTalkNow broadcast. When asked about the state-of-the-art in Intelligent Buildings, he explained, "You need a platform to be able to make corrective changes to the DDC code before you have data streams useful for Machine Learning and AI. Sandstar is that platform." The ultimate goal of Project Sandstar is making it easier for manufacturers to develop new, advanced, cost-effective BMS and IoT devices that work seamlessly in a metadata tag-based environment.

This may all seem very new, but improving our control languages has been on the industry agenda since the early days of DDC. This blast from the past from 2011: “The Past and Future of Control Languages”  speaks to this subject and involves most of the same players. (I find that it is very interesting that the above commentary is likely just the next chapter of this piece of history.)

In this article from 2008, Future of Building Automation, Alper shares his views. Key quote: 

The BAS industry has seen vast improvements in the last 8 years. Protocol wars have ended during the last couple of years.  Integration became an important factor.  It seems that a majority of the buildings are web-enabled.  We have more tools to analyze data and we can communicate better with our buildings.  What does the future hold for the building automation industry?  Before finding out the answers, let’s learn about two new emerging technologies which will revolutionize the building automation industry.

All this history by these industry pioneers leads us to the edge and our meta-morphing mashup.

Of course, we need open edge hardware and open edge software to make this meta-morphing happen. There are currently over six manufacturers supporting Sedona in addition to Alper's company https://ankalabs.com.

Read this interview with Zach Netsov, Product Specialist, Contemporary Controls, Turning a Micro PC into a Controller on how it has happened on the popular, inexpensive micro PC Raspberry PI. 

Netsov:  We're huge proponents of the concept of "open control" and built our BAScontrol Series on the idea of open software and free programming tools. The BASpi was the next logical extension of that idea. We started tossing around this idea of making a controller on a micro PC to give home enthusiasts, students, and DIYers a truly open controller they can easily set up and use. If we can have open software, why not open hardware? We’re fans of the Raspberry Pi and we noticed there weren’t any Raspberry Pi daughter boards which support building automation control requirements, specifically 0-10V analog inputs and thermistor inputs. We thought it would be good to make a daughterboard which can be used in building automation systems. Because we had this technology in our BAScontrol products already, we were able to quickly bring this to a Raspberry Pi daughterboard.

Sinclair:  How does a novice get a BASpi and figure out what to do with it?

Netsov:  We’ve updated the BAScontrol Toolset to support the BASpi. The BAScontrol Toolset includes our Sedona Application Editor, our BASemulator, and our BASbackup tools – and it’s available for free. That means BASpi users can download the free BAScontrol Toolset and use it with the BASpi-IO board or BASpi-SYS to teach themselves sequences of operation based on industry-wide concepts such as graphical wiresheet applications – Sedona, and networked automation using the worldwide-accepted industry standard – BACnet. We love Sedona because it’s a drag-and-drop graphical component programming language that’s open, unrestricted and much easier to use than Python or other procedural languages normally used on the Raspberry Pi. BACnet is very powerful for networked control – it gives experienced building automation BASpi users a piece of mind, and to those who are new to the automation world, an insight of what the professionals are doing out in the field. With the BASpi-IO, we have given Raspberry Pi users the ability to learn and implement their own networked control applications.

Our Building Automation Industry has been thinking about this for a while. I believe that this mashup and new meta-morphing thinking will revolutionize how other industries will connect their standards, their open movements, and expose them to the power of AI. I believe we are on the threshold of an evolutionary forward leap that will make all that we have machine-readable, speeding our transition to a self-learning, and a self-creating world.