Forsler just launched 🌱. Sign up here →

🌳 AI in forestry: Can computers identify trees better than foresters?

Jens Isbak

Welcome to this week's newsletter 👋🏻

February is here, and despite the cold weather, there's plenty of exciting developments in forestry to warm us up! This week, we're exploring the intersection of technology and forestry, from AI-powered tree identification to innovative planting machines and wildlife protection.

💬 In this week's issue:

Foresting Tomorrow #18: AI tree identification and revolutionary planting machines 🤖
Feature story: Inside SmartForest - Norway's push for AI-driven forestry 🔍
Success story: Protecting wildlife with thermal imaging technology 🦌
Product update: Enhanced task summaries in Forsler 📊
Further reading: Sources and insights 🤓

Let's dive in! 🌳

‍

🎙️ Podcast episode 18: AI meets forestry

From AI tree identification to wildlife protection with thermal drones

In this week's episode, we explore a fascinating European study involving 20,000 trees from 33 species. The research tests AI's ability to identify trees using laser scanners and photogrammetry, achieving an impressive 80% accuracy rate.

We also discuss an exciting development in planting technology: a Swedish innovation from Plantma Forestry. Their two-person-operated forwarder with a mounted cabin can plant an incredible 3,000 seedlings per hour, with survival rates reaching 90%! And last but not least we couldn’t avoid speaking of the newly crowned world champions of Handball 🏆

🎧 Listen to the latest episode of Foresting Tomorrow here

‍

📊 Feature story: Inside SmartForest - Norway's push for AI-driven forestry

One of the most fascinating developments in forestry right now is happening in Norway, where the SmartForest initiative is pushing the boundaries of AI-driven forest management.

In our featured article, we explore how image-based AI models are surprisingly outperforming LiDAR in tree species identification, achieving 79.5% accuracy compared to LiDAR's 75.6%. We also look at some interesting limitations - for instance, AI performs much better with trees over 8 metres tall than with smaller ones.

But the key question remains: do we really need AI to identify tree species? While foresters can usually identify species easily, AI might prove valuable for large-scale inventories and monitoring rare or invasive species.

📄 Read the full article on AI in forestry here

‍

🏟️ Success story: Technology saving wildlife

Thermal drone technology is becoming more common in forestry, not just for mapping and planning but also for protecting wildlife. During a recent windthrow clearance, Euroforest Silviculture used a thermal drone to scan the area before bringing in heavy machinery.

The scan revealed a mother deer and her fawn hiding among the fallen trees. Because they were spotted in advance, the team was able to safely move them before starting their work.

This is yet another example of how technology is helping forestry operations work more efficiently while reducing harm to wildlife.

🎥 Watch the whole video here

‍