The Internet of Things (IoT) is finding its way into just about every corner of tactical business operations, including farming. From the AgTech angle, much of what exists on the market today is entirely focused on precision agriculture for traditional (rural) farming. That makes sense, since this is where the biggest ‘smart’ agricultural enterprises ply their trade, to a tune of $6.5 billion a year and growing.
To get a sense of scale, these large operations are spread across tens to hundreds, or even thousands of acres. Many are non-contiguous fields, which means a substantial amount of time, energy, and cost goes into understanding what’s happening in all corners of the operation.
What might not be readily apparent, is that urban agriculture operations, and specifically commercial operations, can themselves be between 1-2 acres on the larger rooftops.
Spread out at the full scale of a city, and you’ve got non-contiguous operations that mirror rural farms, just sitting a few stories off the ground. When you’re talking scale in terms of acres (rural or urban), it makes sense to incorporate some automation and insights into your daily production routine. And that’s where IoT shines.
What is IoT, exactly?
The growing ubiquity and range of the IoT ecosystem makes it a cost-effective approach to capturing distributed information. IoT covers a stunning array of opportunities for embedded sensors to capture and transmit data - all kinds of data. Everything from air quality sensors measuring harmful pollutants to Radio Frequency Identification (RFID) tags geolocating a logistics company’s fleet of vehicles.
Great, but what’s the value?
Harmonized communication between different sensors (“data points”) offers time-sensitive insights that can maximize efficiency and sustainability across operations.
→ In the case of the air sensors, the ability to accurately delineate perimeters of dangerous air quality conditions can lead to citizen alerts on a more directed scale. All cell phones within range receive an alert notifying them of the threat, the threat level, and a suggested evacuation route to leave the area.
→ For fleet asset tracking, fuel usage and optimized delivery routing can save time and cost under normal operating conditions; or, in case of an emergency, geolocation of available vehicles and repurposing to transport people away from danger or deliver supplies into hard hit areas, makes a crucial difference when timing is critical.
These sensor monitoring ecosystems can manage inputs from hundreds or even thousands of devices deployed around agriculture fields or throughout cities, capturing data from multiple vantage points. And with low latency, that data can be viewed securely in near real-time, giving cities a view into their living and breathing metabolic pathways.
More than information, IoT gives us the power to plan
To DirtSat, farming in urban areas carries many of the same burdens and struggles that rural farms endure. Distributed fields require knowledge of local nuances, so adding data inputs can be valuable in changing the way crops are managed.
Enabled by IoT, micro-level insight into an entire operation can help build a new way of harnessing data for peak efficiency.
We’re validating the value of data for urban producers, providing the extra hand to help boost their natural intuition. There’s so much each farmer knows about their growing media, seasonal weather patterns, pest issues, etc., and IoT offers some pretty impressive insights that kick that knowledge up a few notches.
Take water as an example. IoT sensors can both see below the surface, measuring soil moisture / temp and see above ground, measuring even the smallest hint of leaf transpiration (how much moisture is evaporating into the atmosphere from the plants), which says a lot about irrigation quantity and timing. It also forms part of the data chain needed to quantify carbon capture - a topic for another newsletter.
You can start to see the direct connection to cost savings when considering how ‘smart’ irrigation informs decisions. But that’s just the tip of the iceberg. When you add additional tools like remote sensing, you begin to ‘see’ changes in crop health through near-infrared (NIR) light frequencies with the Normalized Difference Vegetation Index (NDVI) algorithm applied. Let’s unpack that.
What can the Electromagnetic spectrum and NDVI tell us?
With geospatial imagery (optical remote sensing), a number of machine learning techniques can be applied to derive granular level crop information across large field areas. Based on the observation of wavelengths and frequencies of light energy the resulting data can indicate anything from crop health to yield forecasts.
“All things on Earth reflect, absorb, or transmit energy, the amount of which varies by wavelength. Everything on Earth has a unique spectral “fingerprint,” just as your fingerprint is unique to you.” (NASA, 2021)
It sounds complex but the NDVI algorithm shows us the amount of light, at certain frequencies in the electromagnetic spectrum, that a plant reflects, which indicates the state of its health (its spectral signature).
A healthy plant will reflect more NIR and green light compared to other bands in the spectrum. The reverse is true for unhealthy plants. When dehydrated or diseased, the cellulose in leaf structure becomes less dense, the plant absorbs more NIR and red light rather than reflecting it.
We can map fields using satellite data and that nifty NDVI algorithm to determine what areas of a field need more irrigation, more fertilizer, or even a fungicide to fend off infection. It looks something like this:
Similar to a traffic light, green highlights healthy crops, yellow indicates an area that may need attention in the near future or it may be transitioning from one state to another, and red signals crops in need of immediate attention.
Time spent on site or tending to crops can be significantly reduced when you know which corners to focus on. Resources can be targeted to the areas in need, instead of being wasted on blanket assumptions about the whole field.
In the urban network of farms that DirtSat is creating, time is a critical resource. Having IoT field sensor data, combined with geospatial intelligence, means non-contiguous fields — at any elevation — can be monitored from anywhere. It doesn’t diminish the need for ground truthing derived from eyes on each rooftop, but a more holistic understanding of crop health comes into focus.
Problems are revealed before they become irreversible.
With IoT and geospatial data, you know exactly what you’re dealing with each day, before you step foot onto a rooftop farm. Better data means better planning, and better planning means better resource utilization.
Better resource utilization means a cleaner, healthier urban environment.
Finding this newsletter valuable? Consider sharing it with friends. You can follow me on Twitter here. Feel free to leave a comment or suggestion below and thanks for being here!