Dawn breaks over a cave system in ancient South Africa. Through blurry morning eyes, you spy the dying embers of a fire lit the night before during an unseasonable rainstorm that has lasted several days. Your stomach rumbles– the storm has sent most animals for shelter, and as a result, you have not eaten in some days. You search for any fruit or nuts you may have scavenged and stored, but you find nothing. Just then, the fluttering of enormous wings just outside the cave entrance catches your attention. You spy a giant eagle having just landed to investigate your camp and quickly run to find your trusty spear. You grab it, but as you turn to the cave entrance, you realize that the eagle has already flown away. You curse your luck and prepare for a hard day of foraging (or, perhaps not knowing the eagle’s intentions, you count yourself lucky).
Since the dawn of human existence, birds have represented an outsized challenge to anyone hoping to track their movements. They embody the ultimate freedom of movement–able to cover miles at a moment’s notice–and, as a result, can be extremely difficult to track. We have come a long way from the days when tracking the movements of birds might be integral to our survival, but we still have a great scientific interest in cataloging their migratory habits. But in these times, from news to commerce and even essential government services, life increasingly happens online. And so (purposely setting aside the fringe conspiracy theory that Birds Aren’t Real), we will explore how one might connect a bird to the internet.
It’s All About Weight
Why can birds fly? The answer is obvious, of course: they are light and aerodynamic. Any IoT solution that hopes to track birds (excepting the flightless varieties) will need to be supremely lightweight so as not to interfere with a bird’s natural ability to soar.
Unfortunately, when it comes to portable IoT devices, there is a constant conflict between three competing concerns: size and weight, battery life, and reporting frequency. Almost all components required to create a functioning IoT locator device are very light and small. GPS, Bluetooth, Wi-Fi, and cellular chips can fit in a smaller space than the average cellular phone. Although technology is continually improving, batteries are still one of the largest sources of added weight in portable electronics. In general, the larger a battery, the more mAh (milliamp hours) it can hold, and thus, the longer it can function, drawing a given amount of energy.
Having already established weight as a primary concern in our bird tracking system, we are limited to a battery that will weigh at most a few grams and a device that reports using a cellular network or other wide area network solution. Locators like this are commercially available, but whether purchasing off the shelf or designing a device on our own, it will be important that our device is properly configured to report in an energy-efficient way. Possible optimizations here are to have the report on an infrequent basis or to include an accelerometer in the device to only report after motion is initially detected followed by a fixed period (in other words, only reporting a position after a bird has flown and landed again in a new location). The first approach might not give us the granularity that we need in the case of birds that move relatively frequently during a day but tend to wind up around the same place each day at the time the device reports. The second approach could lead to premature battery depletion in the case of a highly active bird.
One outside-the-box approach to this issue might be to create a sort of mesh network of birds, similar to Amazon’s Sidewalk or Apple’s Find My Network. In this scenario, a custom-designed bird locator device could use a lower power communication protocol, like Bluetooth. The devices themselves would not be able to communicate directly with the internet. Instead, they would rely on network users having a second device, like a smartphone, with an app installed and always scanning for bird locator tags. If one of these users’ devices detects a bird locator tag, it could then report the tag’s location over the internet to a trusted IoT platform. This solution, unfortunately, falls flat in more rural areas where fewer users in a larger geographic area decrease the chances of a user’s device spying a bird locator tag.
More conventionally, though, we might draw inspiration from the Utah Division of Wildlife Resources’s PeliTrack project, which uses solar-powered GPS backpacks to monitor the long-term migratory patterns of pelicans throughout the American west. Using a solar-powered solution obviates the need for a heavy battery and can extend the lifespan of a locator almost indefinitely—a near-perfect solution.
Without the full financial backing of the state of Utah, we may be hard-pressed to afford state-of-the-art solar-powered equipment. And barring some miraculous new battery technology, we may be stuck observing our birds two days at a time as power runs out on each of our units. Is there another option for us? Yes! And it is a solution as old as birding itself: banding.
You may be saying to yourself, “This is the Internet of Things we’re talking about, how can a metal band with no electronic components fit in here?”. Well, IoT is about connecting the real world to the online world and at the end of the day, even a low-tech solution can provide that gateway. With a straightforward smartphone app and perhaps the addition of a small QR code to each bird band as a quality of life improvement for volunteer researchers, we can bring the practice of bird-banding to the world of the internet. This approach would involve humanely netting birds in a target area, then banding or scanning and recording band QR codes as each bird is gathered and re-released. While admittedly more labor-intensive, this approach has been used to observe birds for over 100 years in the US. Though a low-tech solution on the surface, the quantity and quality of the data gathered would be enormous, mainly when that data is piped to a capable IoT stack with robust data-analysis capabilities.