Ask the Expert: "Why does my asset always appear offline?" - Thing Presence
Meet Anthony. Anthony came to PTC from a large industrial company as a user of Axeda, a leading device connectivity company that PTC acquired in 2014. With a background in aerospace engineering and experience in a variety of industries including life safety, healthcare, nuclear power, and oil and gas, Anthony has been working to create new value around innovation for customers transitioning to ThingWorx. When he’s not working on IIoT, he’s playing music on Cape Cod, photographing Hawaiian landscapes or bringing awesome inflatable chairs into the office.
Kaya: What was the challenge users were facing that led us to create Thing Presence?
Anthony: Axeda customers transitioning to ThingWorx were struggling with the connectivity use case and kept asking, ‘why is my asset always offline?’ When we evaluated it, we discovered that it was tied to the difference between AlwaysOn and Axeda eMessage protocol architectures. IsConnected will always report false for polling and duty cycle devices.
The use case of Thing Presence is to know that the asset is reporting into the network and is ready to provide information (push) or be accessed to retrieve information or do a remote desktop support (pull). This use case is relevant for any asset in ThingWorx that uses duty cycle.
In ThingWorx 8.4 (coming in early 2019), the new IsReporting state will inform the user when a polling device is communicating on a regular basis. If it is, then IsReporting will be true. The IsReporting state resolves the discrepancy wherein devices that are on duty cycle appear disconnected due to the IsConnected state reporting false. New "IsReporting" state improves visibility of an asset's communication state
Kaya: How exactly does Thing Presence work?
Anthony: You can think of it in terms of having teenagers. You tell them they need to check in with you on a regular basis through text message. If a text is missed, all of a sudden you take action.
Now, imagine the teenager is a device. If a device was supposed to check in every five minutes and it misses one poll, I want to flag that as a problem. The challenge with that from a service perspective is that sometimes your service personnel will go out and work on a device and may need to take it offline for a bit of time; we need to factor that in. We certainly don’t want to deploy someone or try to fix something when a service technician is already there.
You might decide, ‘my average service visit is an hour, so, if I miss a couple of pings, I’m okay; but, if I’m offline for more than an hour, then I’d like to know about it because I’d like to take action.’ Thing Presence allows you to define that window.
Kaya: You’ve mentioned ‘duty cycle’ and ‘polling cycle.’ Can you explain these terms?
Anthony: Duty cycle and polling cycle are the same thing. It means that a device has a time for which it is expected to check in, and, provided that it checks in within that timeframe, all is good with the connection.
Connected services rely on a connection. As soon as the connection is broken, I no longer have the ability to service the asset.
Kaya: Given everything we’ve discussed, where do you see Thing Presence headed?
Anthony: The next piece of the equation for us is to provide information on the health of the connection. When you look at servicing a remote asset, you need to a) know that it is communicating, and b) know that the connection is healthy before you try anything. I wouldn’t want to try a software update if I am losing connection with my asset on a regular basis.
What do we mean by health? We mean: is the device checking in when it should be? If it’s not, is there a pattern to that connection, and are those patterns tied to applications? For example, is it only on during working hours? Does it turn off during holidays? If the device is in a school, does it turn off during summer maintenance work? This allows us to garner insights on how and when the equipment is being used, not just operating status. At the end of the day, does this mean I can apply analytics and AI tools to it? Absolutely. Is it the first place I would apply it? Probably not.
Kaya: In your mind, what is the next big thing coming in ThingWorx that you’re particularly excited about?
Anthony: Mashup 2.0 and Asset Advisor 8.4. (Double mic drop.)
Kaya: That’s awesome. My last question is related to you. Can you tell me what your favorite aspect is about working at PTC?
Anthony: The chaos. Very often it’s chaos that breeds innovation. What I mean by that is that, if you try to create something because you sit down and you say, ‘I am going to innovate,’ very often that is a failure because the majority of the time it’s the influence of a deadline, a customer need or an application at hand that makes the environment trying and sometimes hectic. But, it is in these challenging environments where you can be the most creative and innovative as an engineer.