The old ways are changing. Technology moves fast, and with it our methods and strategies for managing the landscape. Artificial Intelligence, or machine learning, presents an incredible opportunity in forest management. Better and faster data means more time spent on other values. Values like biodiversity, clean water, carbon storage and recreation. Today we find out how far we have come and what lies ahead! Hopefully not Skynet…
Resources
Sponsors
Takeaways
Technology in forestry (3.53)
Rolf is the CEO of Collective Crunch while Christof is the CTO. Their mission in co-founding the company was to use advanced analytics with environmental data to improve business processes. They both had a connection to nature growing up, which has continued in their work today. Forests are to Finland what cars are to Germany, so it was the perfect fit to establish a tech startup for forests there.
Forest inventories (13.22)
Christof notes that forest inventories came into being after the overuse of forest resources in the industrial revolution and the realization that those resources are not unlimited. They need to be managed by noting how much there is and how much can be extracted. Country-wide inventories began only 100 years ago, and he claims Nordic and central European countries were the forerunners in this matter. Smaller-scale inventories are done through statistical sampling, while private land inventories are not publicly available.
Forestry is an area where AI actually works (19.07)
Christof points out that LIDAR is done in 5-10 year cycles since acquiring data is expensive. They enrich data using their software to fill the gaps, making it cheaper. Rolf believes that using data to efficiently use commercial forests is in the interest of society because it leaves room for natural forests. He lists some of the areas their company is considering branching out into and shares an example of how data can help track biodiversity loss to keep corporates accountable for their processes and offer transparency of ethics to consumers.
Data-driven baselining (27.52)
Rolf discusses one of their services, data-driven baselining for project management and bringing data into greenwashing conversations. Christof highlights that data can be processed at different scales, depending on the needs of the project and if acquiring that scale of data would add value to the project. He adds that AI helps with squeezing the most out of the available data in a project. Rolf recognizes the benefits of this for a large landmass managed by a single entity.
The data you need (34.08)
Christof lists the various tools to acquire data at various levels of granularity and mentions that their customer segment usually looks for the medium scale of data. He talks about how modern AI systems are based on extracting and learning from existing data, the relationships between the data and what you want to predict. The quality of the data they extract from the forest makes all the difference.
Limits (39.24)
Rolf finds that the kind of data available can be limiting while Christof says ground measurements can be a limiting factor. In markets where ground truthing doesn’t exist, Rolf is optimistic that they can design statistically valid and plausible ways to make predictions. Where LIDAR exists, technology can be used to augment the data between flights, and he believes using two technologies to derive two data sets is a great way to have technologies complement each other.
Valid and reliable (51.19)
Rolf talks about how data can provide a verifiable and scientific way to manage forests, and they are trying to convince the industry to become data-driven for scale. Their company can provide baseline analysis to help investors who are looking to invest in carbon credits. The policy around carbon offsetting is complex, and its efficiency is measured differently in every country, Christof says, further complicating the additionality claims of projects.
The North American story (1.00.01)
Christof points out that even though restoring Canadian forests destroyed by fire would not be economically attractive, it should be done nevertheless. Rolf foresees the carbon price continuing to go up in North America, which may incentivize forest owners to leave forests unharvested. Christof has observed change management initiatives in government forest organizations to embrace the new ways of managing forests.
The data-driven future (1.10.08)
Christof believes more and higher quality data is a matter of time. They are seeing exponential growth in data, and his dream is to have a digital equivalent of the world’s forest cover, which would expand the possibilities of forest planning and management for the most economically optimal operations. Rolf adds that higher-resolution data is needed for biodiversity initiatives, but that we are on our way there. He encourages the application of the data that exists today instead of being skeptical.
If you liked this podcast, please rate and review it, share it on Instagram, Twitter and Facebook and tag a friend, and send your feedback and comments to yourforestpodcast@gmail.com.