Clouds on the ground: the potential of digitalized trees

Kyle Nielsen, our innovation consultant, has written several blogs about model inputs and elaborated on the ground truth process. In this article, he elaborates on a few technologies Acorn has been testing over the past year and imagines future technological possibilities down the line.

Acorn has always been at the forefront of the carbon market by leading the industry in the creation and certification of scalable remote sensing-based carbon measurements. The main aim is to reliably measure the increase in biomass of our trees every year. As carbon markets have come under more scrutiny recently, I’ve been thinking about how can we bring people into the field and actually see the work on the ground.

Working with Agerpoint’s technology

We’ve been testing upcoming technologies with Agerpoint. As in our current methodology, we use airborne LiDAR to correct our ground truthing. The airborne LiDAR measures distances from the source to the ground extremely accurately and we create point cloud observations to reliably measure the height of a tree. But imagine if we could also create point clouds on the ground.

We’ve been trying just that with Agerpoint. They use iPhones’ built-in LiDAR technology to create useful observations on the ground. By walking around the plot whilst the phone takes hundreds of images, you can use Agerpoint to combine these images to create a digital model of the trees.

The added value of digital 3D trees

A digital 3D tree is cool, but is it useful? Absolutely. On the ground, there are always branches hitting you in the face, fallen trees and paths to navigate, leaf litter to make your way through, and so on. But with Agerpoint, after you’ve walked around the farm, they can automatically identify the tree trunk. For that tree trunk we then get a specific GPS point per tree, and the diameter at 1.3 meters high for all the trunks.

Now, if we want to return in later years to see the tree growth, we have the dimensions and the specific tree’s GPS so we don’t get confused which tree was which (more difficult than you think). Additionally, once we have the GPS per tree and on-the-ground measurements, we can link that data to the airborne LiDAR and create a full model of the tree. And if I can already recognise the tree species from an image, surely a computer can as well.

In essence, Agerpoint offers a powerful digital measuring and reporting tool that’s straightforward to use and easy to carry around in a pocket.

"Agerpoint is proud to partner with Acorn as we build the future of accessible, accurate ground truth. As digital tools in the field augment airborne measurements, AI-powered insights will guide farmers and industry leaders to smarter decisions."

—Kevin Lang, CEO, Agerpoint

Limitations and ambitions

While promising, this is all still something we’re testing — as certifiers have yet to accredit this technology for use, and a major limitation we’re finding is the lack of localised data. There are a few open-source tools such as Google Lens and similar companies that claim to be able to identify tree species, but we are finding a lack of accuracy for local African and Latin American species, as many of the companies and databases are trained on mainly European and American data sources.

Focusing on the future: if we can continuously collect these tree models, we could see how certain trees are growing in certain areas. Everyone is talking about AI these days, but where I see a huge benefit of AI is localised insights that are often left out for smallholder farmers. One of the most difficult parts of the Acorn process is the agroforestry design itself: the direction of the sun, rainfall, prevailing wind, soil type, climate, relief, farmers’ preference, fruit or non-fruit-bearing trees, and so on.

In the future, using the insights I’ve discussed above, farmers could get feedback on what agroforestry designs would be suited for their area and get answers to their questions about prices, weather, disease, et cetera. Or if certifiers and buyers would like to visit the farms and actually see what’s on it, they’d be able to do so digitally.

The future is limitless, but I’m told I do have a maximum word count, so I’ll leave it at this for now. Stay tuned!