Memento Mori: When Will My IoT Devices Die?

The lifetime of a wireless Interet of Things (IoT) device depends on:

• How much energy is being used
• How much energy is available

In this article we focus on the first of these two: how is energy being used. With the Thingsquare system, we keep track of the energy usage, and use this to estimate the lifetime of every device, in the field and during development.

For most IoT devices, the amount of energy that is available is determined by the size of the battery. (There are other ways to power IoT devices, which we cover in another article.)

What consumes power?

For wireless IoT devices, the energy is spent by several different mechanisms:

• The radio, in listen mode
• The radio, in transmit mode
• The microprocessor (CPU)
• External chips, such as sensors
• Any idle energy draw

The list above is ordered with the most energy consuming mechanism at the top, and the least consuming mechanism at the bottom.

The power consumption data from the LPSTK hardware can be seen below:

The current draw of the CC1352R chip, from the data sheet: CPU (left) vs radio (right). Note that the CPU numbers are measured in micro-Amperes whereas the RF numbers are in milli-Amperes – a thousand times higher.

The idle energy draw comes from the power needed to keep the microprocessor in its lowest power mode, keeping things like memories in standby mode, as well as any leakage in the hardware. In the final hardware design, those are typically reduced to a minimum.

The radio must be carefully managed to keep its power consumption down. The Thingsquare IoT platform uses several mechanisms to manage the radio duty cycle. Some of the parameters can be tweaked, but otherwise it will mostly manage itself.

Keeping track of the power consumption

The Thingsquare system keeps track of the power consumption of each device by measuring the time it spends with the CPU and the radio turned on and off, respectively. This data is periodically reported to the backend.

The benefit of using software to keep track of the power consumption, as opposed to using hardware to measure it, allow us to both track the power consumption at a very large scale and over very long time. We can easily track the power consumption of thousands of device in the field, over several years. This is not possible to do with hardware-based methods.

Based on the reported power consumption, we can estimate the lifetime for each device. This lifetime depends on the size of the batteries that are being used.

How to see the power and lifetime

By default, devices periodically report their power consumption statistics to the Thingsquare backend. To get a good amount of data, let the devices go about their daily lives for a few days.

Having a few days of data will also allow seeing trends in the data. For example, devices may be more active during the day. This will be evident from the statistics.

To find the power statistics and lifetime estimates, we use the Thingsquare app or the equivalent Thingsquare web frontend.

To see the statistics for a given device, find that device in the device list and tap on it.

Next, tap the Details tab on the lower right. When on the Details view, find the Lifetime Estimate button on the right hand side of the screen and tap it.

The Lifetime Estimate view brings up a long list of details on the power consumption:

• The overall estimate of the total current consumption over the last days
• The lifetime estimate for three different battery types: CR2032 (coin cells), AAA, and AA
• The breakdown of where the power is spent: CPU, mesh/sub-GHz radio listen, mesh/sub-GHz radio transmit, Bluetooth radio, given both as percentage of the total time and as current estimates

Scrolling down in the view shows the breakdown over time as a graph. This allows seeing trends in the data over the days of measurement.

Tapping in the graph shows the breakdown of that particular point in time.

For the LPSTK, which uses two AAA batteries, the lifetime estimate is the AAA battery value, but doubled.

Conclusions

The lifetime of a wireless IoT device is determined by how it spends its energy and how much energy is available. The Thingsquare system keeps track of how much power is being consumed by every device in the system. This allow us to estimate the lifetime of a device.

To see the power consumption and lifetime estimate, let the device be online and report its data for a few days. Then use the Thingsquare app and tap on a device in the list, go to the Details, and tap the Lifetime Estimate button.

Interested in using this in your own IoT projects or products? Get in touch with us and we’ll be happy to schedule a demo!