Monitor

As far as data density goes this is the page that is the most stacked. It displays all data from onboard sensors, the state of actuators, etc. If we had to list them they would fall into the following categories.

• State - target temperature, open/closed state,

• Sensor measurements - temperature, humidity, CO2, LUX, PIR state

• Battery - the voltage level at the last uplink

• Network - RSSI, SF, Frame Count

The aforementioned give a good indication of the current state of the device, however they only provide data from the last keep-alive uplink. The real benefit of MClimate devices and their use with Enterprise comes from looking at data over time and adjusting the control parameters accordingly.

It is vital one understands how parameters change over time and what insights can be extracted from this variations to better understand how the environment in a building changes. Thus, let us look at some example data.

Vicki graphical data example

We are going to look into some data from a Vicki Smart Wireless Thermostat over the course of a day to better understand its potential uses.

As one can see the graph gets a bit overloaded with different curves, thus it is considered a good practice to look at two separate visualization use cases. You can select which parameter to display/hide on the graph by clicking on its name below the graph (they are all color coded).

• Main parameters graph

• Network parameters and battery graph

Main parameters graph

This view is very much what you want to be looking most of the time as it pertains to information that directly impacts the environment in your facility both in absolute and perceived terms

The parameter displayed in this use case are Target and Measured Temperature, Humidity and Valve opening.

As it can be seen the Measured temperature is very much higher compared to the target temperature. This is normal in this case as the readings are taken in the summer and there is Vicki has no cooling capabilities. What is important is that since the Target temperature is a lot lower than the Measured, the Valve opening is set to 0% in order for the valve to close completely and limit the flow of water, which is the proper behavior. The system, thus is reacting as it should be.

The blue curve is the Humidity one, its importance is often overlooked however it might be indicative of environmental issues. It is considered the norm for an indoor environment to have 30%-50% relative humidity. This ensures that one's perception of the indoor climate is not shifted due the air being more humid for example (this makes one feel warmer, especially when the temperature is also higher). You want it to be around 40% as is in the graph, ideally. In this case it is way too high for it to be comfortable as it is very humid and warm at the same time (this also increases the risk of mold forming).

In this example the target temperature remains constant, and since the ambient summer temperature is high the valve remains constantly closed, however it would be normal if there is a sharp change in the Target temperature that there is an equal response in the Valve opening.

For example if one would change the Target temperature from 15 to 25 the Vicki will open the valve 100% to rapidly heat the space (where it might have been entirely closed to keep the temperature low at 15).

The Valve opening curve is useful in cases where troubleshooting is required, for example if there is Target temperature change and there is no change in the Valve opening percentage. This indicates an issue with the valve itself or perhaps the Vicki.

To sum it up, you want the curves to be smooth to have a consistent environment that feels well to be in. In case there is a sharp change in the Target temperature, the other curves should follow, the Vicki reacting by adjusting opening and the Measured temperature following in response. Any other behavior is indicative of an issue.

Network parameters and battery graph

We consider this view a supplementary one that gives you insights into the performance of the device on a LoRaWAN network level and its battery life (the two are related).

This view is mostly useful if you want to gain insight into the operation of the device on a network level. It is often the case that one would look at these parameters if an issue has presented itself in the environment (either in absolute metrics or perceived quality). You want the parameters curves to be smooth as in the example image above (with the exception of the RSSI, which is expected to vary due to the nature of RF wave propagation itself).

If one were to look at this view for the purpose of troubleshooting, we would recommend to look into the curves in the following order:

• Frame count - this should always be linearly increasing. Every subsequent keep-alive increases the Frame count by 1, thus if there is a jump in the value it indicates data has been lost, which could indicate poor network coverage. too much interference due to too many devices being connected to a single gateway, etc. If the count reset and starts from 1, this means that Vicki rejoined the network, which could indicate tampering, power issues or a gateway being disconnected for a time. In any case it warrants investigation.

• SF - the spreading factor is an indication of the network coverage/quality and whether efficient communication is present. Ideally you want SF7 to be the most often used one, with SF8 and SF9 being used as little as possible. SF10 to SF12 are indication of a poor signal quality/network coverage, we do not recommend operating a network of devices under those conditions. Look into adding more gateways in this case. What is important to is that SF directly impacts battery life. For example, a Vicki working with its default settings for an 80 days heating period would have 12+ years of battery life on SF7 and around 3 years on SF12, resulting in a four-fold lifetime decrease.

• RSSI - this supplements the SF curve. When the RSSI drops sufficiently low, the LNS compensates (if ADR is enabled) and goes on a higher SF, which improves connection stability but hurts battery life and creates potential network congestions. If you are mostly working on SF7 to SF9 you should not be worried too much about the RSSI, however if SF degrades, look into the RSSI levels and try to correlate with other parameters (gateways dropping, suboptimal location near metal construction, etc.)

• Battery - we now know that the other parameters we previously discussed all affect battery life and operating under suboptimal conditions drains the battery quicker than desired. If you are operating an efficient network, utilizing good batteries (MClimate devices come with Energizer L91, which are the highest quality on the market) and your devices are transmitting within a reasonable keep-alive period (10 minutes for example) you should see little to no change in this curve, as it will take years for the battery to drain significantly. Do note though that once the battery voltage reaches 2.8V, you should look into replacing it as the 2.7V is the threshold for normal valve operation (the ability of Vicki to exert sufficient force to close a valve).

Let us end with a final example that indicates a network issues.

It is immediately noticeable that the Frame counter started anew, meaning that the Vicki was forced to rejoin the network. It is very easy to identify the event and it would immediately bring this to the attention of a maintenance person for example, giving them information so they can evaluate if they should investigate the issue further.