… hopefully not literally!
When planning IoT projects, there’s often one part that tends to get overlooked—power.
The traditional M2M and telematics world was largely based on single-purpose sensors, e.g. a water meter reporting flow rate once a day. These devices would wake up, send their data, and then power down again. However, as the IoT continues to permeate through more market sectors, solution providers are finding new remote use cases where the power draw is more of a concern than ever.
Power draw comes out of a few things:
The CPU of the device
More complex CPUs such as SoC (system on chip) style units require more power. For example, the Raspberry Pi 3 is recommended to run on 2.5 Amps of power—more than a typical phone fast charger!
The sensor gathering the data
Power for this is usually minimal
The data connectivity method
Connecting to a cellular network requires not only more CPU cycles, but more power to transmit, receive, handshake, connect, disconnect, and send its data, versus a sigfox or LoRA type device).
The feature creep threat
As IoT evolves, feature creep evolves with it—and the more the device has to do, the more power it requires. 10 years ago, a water meter would read flow rates once a day. Fast forward to today, it might be providing a wifi hotspot for the engineer, or perhaps a continuous video feed of the pipe or the immediate environment. All this requires more data, and more power to stay on and transmit that data. In addition to power, connectivity type (LoRa, Sigfox, WiFi, or cellular) should be considered carefully in feature creep situations as some technologies fit better than others.
Powering the Big Data Cycle
On 30th July, our Business Development Manager, Terence, will be taking part in RideLondon, cycling a 100 mile course from London to Surrey. We’ve set up the Big Data Cycle project to track Terence’s training using IoT enabled devices. These devices measure environmental points as well as rider performance, giving us valuable insights into the surrounding area and Terence’s training progress. You can read more about the project in our first story here.
Kitting out his bike with a host of connected devices that monitor pollution, air quality, speed, and location, to name a few, requires a host of micro USB cables and a battery that can support all that power.
Terence being the lovely accommodating person he is, is more than happy to carry all this equipment, but at the same time we don’t want to weigh the bike down too much with devices and multiple power sources inappropriate for the task at hand. The battery and devices need to be dimensioned to suit the purpose. In this case, Terence’s ride will last 4-6 hours (depending on how many Mars bars he has to power himself!) so we’ve put together an IoT solution which roughly has the following criteria:
|Capture data points (i.e. air quality, location)||Small sensor board|
|Provide data points every minute, and video / imagery of the ride from a mounted GoPro on the bike||This works out as approximately 10MB an hour, so we’ll need a SIM with at least 100MB of data (with a bit of wiggle room)|
|Prevent connectivity from dropping out to ensure seamless and accurate data collection||Pangea multi-network SIM cards roam across all four major UK networks, ensuring the best connectivity|
|All equipment needs to be powered for 4-6 hours||The combined power draw of the devices will be 250mAh (milli-amp hours), so we’ll need something which can provide 250mAh for 6 hours. Round this up to 2000mAh, and a small battery pack would suit.|
|Collectively, the equipment should be as lightweight as possible||Based on the requirements above and our calculations, we’ll be using a small set of sensors, a lightweight GoPro, and a lightweight 2000mAh battery (slightly larger than a typical lipstick).|
Start with power
So, what happens when we don’t dimension power requirements correctly?
In short, we put the solution at risk! It’s important to be exact in the specification and be clear about what the solution is to achieve. In Terence’s project, our marketing team were looking to use the GoPro to stream HD video live. While that may have been great in theory, the GoPro has a running time of 1 hour with continuous video, so the battery charging the GoPro would also have to be dimensioned to accommodate that. The battery now goes from lipstick-sized to five times that, with a corresponding increase in weight—which Terence couldn’t be expected to cycle with.
If you want to know more about the Big data cycle, check it out here. If you want to know more about Pangea, multi-network SIM cards, pollution sensors, connected cycling, or just Terence, get in touch with us here.