As we close the final chapter of 2024, let’s take a moment to reflect on the milestones achieved this year and explore fresh inspiration for the year ahead filled with new discoveries, groundbreaking ideas, and continuous learning.

2024 Recap

This is definitely not a complete list of all the interesting topics, but rather a summary of key directions.

To make your reading easier, press Ctrl + F and type what you’re interested in specific crop or topic for example, ‘tomatoes’ or ‘vegetable’ and you’ll find your section immediately.

Let’s start by listing what we’ve learned!

AI for Soil

However, the advancements of AI in soil science are significant, yet soil remains full of untapped opportunities for AI applications (for example, bioremediation). Moreover, due to the complexity of soil microbiome, we won’t uncover all its secrets in the coming year. It will take at least a decade, given the current capabilities of AI.

Soil microbiome-plant interactions. Source: Jansson et al., 2023. Image is publicly provided by Aspen Global Change Institute

*Ternary plot of primary biomarkers Ternary plot of primary biomarkers (left: prokaryotes, right: fungi) associated with various fertility sources. Colors represent the phylum or phylum/family of the biomarkers, and the symbol size indicates its scaled contribution to identifying the fertility source. Source: Mo et al., 2024

AI for Plant Health

We discussed the application of AI to address monocropping challenges and monitoring using a regular camera (2D -> 3D). Plus music to boost crop yields! We talked about crop health monitoring and yield prediction, including examples involving sorghum, grapevine, pitaya (dragon fruit), cannabis, okra, onion, and phenotyping pepper in various environments

*Ternary plot of primary biomarkers Ternary plot of primary biomarkers (left: prokaryotes, right: fungi) associated with various fertility sources. Colors represent the phylum or phylum/family of the biomarkers, and the symbol size indicates its scaled contribution to identifying the fertility source. Source: Mo et al., 2024

Sorghum panicle segmentation. Predictions were performed by the models trained using the frameworks Detectron2 and Yolov8 over sorghum images. Source: Santiago et al., 2024

Classification of ripeness of pitaya fruits. (a) Bud. (b) Immature. (c) Semi-mature. (d) Mature. Source: Qui et al., 2024

Drones? Who said that magic word :)!

Multi-rotor DJI Matrice 600 Pro system equipped with MicaSense RedEdge-MX sensor (a) and a MicaSense Calibration Reflectance Panel serial number: RP04-1918107-OB (b). Source: Nduku et al., 2024

We used them quite a lot in relation to:

• monitoring corn (maize) planting density,

• sugarcane phenotyping,

• wheat height prediction,

• amaranth fertilization,

• detecting cotton top buds,

• apple fruit detection.

Plus we touched a topic of weed detection for canola (rapeseed) crops.

Example plot of Bud-YOLO model detection results. Source: Zhang & Chen, 2024

Comparison of the recognition of overlapping apples using NMS and Soft-NMS algorithm. Source: Ji et al., 2024

Plus (not about drones but about apples) - more news are coming soon based on this post:

AI for Plant Breeding

Speed Breeding is gaining momentum as the demand for climate-resilient and high-yield varieties grows rapidly. Plus crop-focused breeding of peanuts, carrot, sugarbeet, malt barley.

Biofortification is transforming staple crops into nutritional powerhouses. Examples: golden rice and golden lettuce Vertical Plant Factories are trending as the ultimate solution for urban food production (combining AI + robotics + hydroponics). Example: AI for rice breeding plus Breeding for Indoor Vertical Farming

The PF speed-breeding system for rice seeds. Source: Liu et al., 2024

Summary of application of multi omics approach in malting barley. Source: Panahi et al., 2024

AI for Pest Detection

Smart pest detection is proving to be a vital tool in agriculture, from hydroponically grown tomatoes, detecting melon fruit fly, to detecting thrips in mango orchards and bark beetle identification in forestry.

Trapping and detection devices. (a) Outdoor trapping and detection devices. (b) Method of using an attractant. Source: Wei & Zhan, 2024

In hydroponics, AI systems can identify early signs of pest infestations via real-time monitoring pest traps, ensuring timely intervention to protect delicate environments

More great news about AI-powered pest trap monitoring with autonomous camera is coming soon. A quick insight is provided below.

AI for Disease Detection

A huge topic worthy of a book, AI for disease detection in crops is all about advanced image recognition and data analysis. We explored the application of AI for detecting diseases on coffee leaves, fusarium on wheat, citrus greening using UAVs, banana diseases, potato late blight detection. Separately we’ve looked at greenhouse vegetables diseases.

IoT plant-level video stream inferences using YOLOv5-small model. Photo credit: Kontogiannis et al., 2024

Visualization of detection results. Source: Wang & Liu, 2024

AI for Controlled environment agriculture (CEA)

Apart from traditional crops for controlled environment agriculture, like lettuce, strawberries, tomatoes or cucumbers, we also explored vegetables and medicinal plants. Additionally, we took a quick look at urban agriculture, which isn’t always about CEA it can also involve growing on walls and roofs!

Images of strawberries at varying levels of ripeness under different colour spaces. Photo credit: Karki et al., 2024

Controlled-environment plant production systems, also called vertical farms or plant factories. Images (A) and (B) courtesy of the Intravision Group, with permission. Source: Dsouza et al., 2024

In August, I published a book titled Vertical Farming: A Guide for Growing Minds, and there is even more exciting news about it coming soon! My book is available on Amazon, Ebay or just drop me a message and we’ll discuss the way for you to get it.

Read the inspirational post about the “Vertical Farming: A Guide for Growing Minds” here

Robotics in agriculture + Machinery & Automation

We didn’t delve too deeply into robotics this year, touching only on a few use cases on robotics & automation as well as smart machinery. Probably, soybean spraying, Variable Rate Application (VRA) for fertilizers are worth mentioning as well as the so-called ‘fake harvesting robot’. This is a gap we’ll definitely address in 2025

Online multi-sensor platform used for soil data collection. Source: Qaswar, Bustan, Mouazen, 2024

Everything in between

Plus, we have dived into a few other interesting topics:

→ forestry as well as agroforestry, → winemaking, → agrivoltaics, → weather forecasting → smart irrigation

Tree segmentation qualitative results for models, which are trained using the Arbocensus tree segmentation dataset. The worst outcomes are shown in the first three columns, while the best outcomes are shown in the remaining ones. The cyan area represents the true positive pixels, magenta regions are false positive pixels, and yellow are false negative sections. The white solid line represents the ground-truth boundaries. Photo credit: Arevalo-Ramirez et al., 2023

A research project in Baden-Württemberg is testing an agri-PV system with fruit cultivation near Lake Constance. Source: Fraunhofer ISE

Trends for 2025

• Agricultural Robotics: A significant leap forward is expected, with robotics playing a larger role in automating tasks from planting to harvesting.
• Smart Irrigation, Smart Machinery, Smart Yield Mapping: Precision agriculture will see enhanced integration of AI and IoT to optimize water, equipment, and productivity.
• Vertical Farming Growth: This innovative farming method will expand further, driven by advancements in technology and urban demand for fresh produce.
• Regenerative Agriculture: Focused on soil health and sustainability, this approach will gain momentum for its environmental benefits.
• Temperature-Tolerant Crops: With climate change impacting agriculture, the development and adoption of resilient crop varieties will accelerate.
• IoT and Sensors: Increased deployment of connected devices will enable real-time data collection, leading to smarter decision-making on farms.
• Renaissance of (Real-Time) Monitoring: Monitoring tools will evolve, offering instantaneous insights for crop health, pests, and environmental conditions.
• Blockchain in Agriculture: From supply chain transparency to traceability, blockchain will solidify its position as a transformative tool in the agricultural sector.

Thank you for being with us in 2024 and… let’s make 2025 truly agri-awesome together!

Youtube - AI in Agriculture Podcast: Seeing Green in Affordable Plant Phenotyping