60 acast.com Future of Agriculture Podcast https://feeds.acast.com/public/shows/future-of-agriculture en Maryna Kuzmenko ai in agriculture, smart farming,precision agriculture Maryna Kuzmenko Welcome to the Future of Agriculture Podcast – AI Promise and Crop Production Reality. This weekly show highlights the most striking, useful, and educational innovations in agriculture and discusses current AI trends with a farmer-first lens. We cut through hype, translate research into plain English, and turn tools into practical SOPs that improve yield, quality, profitability, and sustainability.


🎙️ What you’ll get every week

• Field Note: a farm or greenhouse problem and how tech actually solved—or failed to solve—it.

• Toolbench: hands-on looks at AI/ML, sensors, drones, robotics, computer vision, LLM copilots, and edge devices.

• Paper-in-Plain-English: new research distilled into operational takeaways for agronomy and management.

• Myth vs Reality: separating marketing from agronomic constraints, data limits, and regulation.

• The 60-Second ROI: quick math on cost, benefit, risk, and break-even.

• Listener Q&A: your challenges, answered with concrete next steps.


🌾 Who this is for

Farmers, agronomists, CEA/greenhouse growers, orchard and vineyard managers, crop advisors, ag retailers, input suppliers, agtech founders, investors, researchers, students, and policy shapers who want to know where AI truly creates value in crop production.


🤖 Topics we cover

Precision agriculture; variable-rate irrigation and fertilization; soil health; satellite and drone imagery; multispectral and hyperspectral sensing; canopy vigor and greenness indices; computer vision phenotyping; plant counting and sizing; fruit load estimation; quality grading and defect detection; disease and pest scouting; early warning; LLMs as agronomy copilots; recordkeeping and traceability; on-device and edge AI; robotics; decision support; compliance; and economics.


🧪 From lab to field

We connect machine learning advances (object detection, self-supervised features, multimodal transformers), statistics, and control to real constraints: sampling bias, lighting, occlusion, phenology, calibration plates, label drift, model drift, edge compute, connectivity, human workflow, and data governance. Expect frank talk about false positives, thresholds, and how to validate models against ground truth.


🥕 Crops and systems featured

Apples, berries, tomatoes, potatoes, grapes, kiwifruit, leafy greens, brassicas, citrus, stone fruit, cocoa, coffee, cassava, maize, wheat, rice, soy, sugar beet, greenhouse vegetables, vertical farms, and regenerative, organic, and conventional systems.


📈 Why subscribe

To save time, money, and frustration. Each episode converts the week’s most useful ideas into takeaways you can test in a field, trial block, or lab. Learn when a simple SOP beats a fancy model, when to rent a drone versus hire a scout, and how to measure success so pilots become deployed systems—not shelfware.


🧰 Workflows we unpack

• Drone or phone: match sensor, altitude, and ground sampling distance to the job.

• Spot-spray: detect → decide → actuate with rate cards and validation plots.

• Fruit count and size: from flowering to packhouse reconciliation.

• Leaf area, greenness, and nitrogen: when SPAD, when RGB, when multispectral.

• Disease triage: early signs, thresholds, and escalation rules.

• Greenhouse setpoints: combine vision, climate, and fertigation to steer vigor.


🗓️ Publishing schedule

New episodes every week. Expect rotating deep dives, farmer conversations, startup case studies, and research spotlights.


⚖️ Our promise

Plain talk, practical evidence, and transparent limitations. We disclose conflicts, avoid vendor hype, and always describe the context, dataset, and validation design behind any result.


Hosted on Acast. See acast.com/privacy for more information.

]]> Welcome to the Future of Agriculture Podcast – AI Promise and Crop Production Reality. This weekly show highlights the most striking, useful, and educational innovations in agriculture and discusses current AI trends with a farmer-first lens. We cut through hype, translate research into plain English, and turn tools into practical SOPs that improve yield, quality, profitability, and sustainability.


🎙️ What you’ll get every week

• Field Note: a farm or greenhouse problem and how tech actually solved—or failed to solve—it.

• Toolbench: hands-on looks at AI/ML, sensors, drones, robotics, computer vision, LLM copilots, and edge devices.

• Paper-in-Plain-English: new research distilled into operational takeaways for agronomy and management.

• Myth vs Reality: separating marketing from agronomic constraints, data limits, and regulation.

• The 60-Second ROI: quick math on cost, benefit, risk, and break-even.

• Listener Q&A: your challenges, answered with concrete next steps.


🌾 Who this is for

Farmers, agronomists, CEA/greenhouse growers, orchard and vineyard managers, crop advisors, ag retailers, input suppliers, agtech founders, investors, researchers, students, and policy shapers who want to know where AI truly creates value in crop production.


🤖 Topics we cover

Precision agriculture; variable-rate irrigation and fertilization; soil health; satellite and drone imagery; multispectral and hyperspectral sensing; canopy vigor and greenness indices; computer vision phenotyping; plant counting and sizing; fruit load estimation; quality grading and defect detection; disease and pest scouting; early warning; LLMs as agronomy copilots; recordkeeping and traceability; on-device and edge AI; robotics; decision support; compliance; and economics.


🧪 From lab to field

We connect machine learning advances (object detection, self-supervised features, multimodal transformers), statistics, and control to real constraints: sampling bias, lighting, occlusion, phenology, calibration plates, label drift, model drift, edge compute, connectivity, human workflow, and data governance. Expect frank talk about false positives, thresholds, and how to validate models against ground truth.


🥕 Crops and systems featured

Apples, berries, tomatoes, potatoes, grapes, kiwifruit, leafy greens, brassicas, citrus, stone fruit, cocoa, coffee, cassava, maize, wheat, rice, soy, sugar beet, greenhouse vegetables, vertical farms, and regenerative, organic, and conventional systems.


📈 Why subscribe

To save time, money, and frustration. Each episode converts the week’s most useful ideas into takeaways you can test in a field, trial block, or lab. Learn when a simple SOP beats a fancy model, when to rent a drone versus hire a scout, and how to measure success so pilots become deployed systems—not shelfware.


🧰 Workflows we unpack

• Drone or phone: match sensor, altitude, and ground sampling distance to the job.

• Spot-spray: detect → decide → actuate with rate cards and validation plots.

• Fruit count and size: from flowering to packhouse reconciliation.

• Leaf area, greenness, and nitrogen: when SPAD, when RGB, when multispectral.

• Disease triage: early signs, thresholds, and escalation rules.

• Greenhouse setpoints: combine vision, climate, and fertigation to steer vigor.


🗓️ Publishing schedule

New episodes every week. Expect rotating deep dives, farmer conversations, startup case studies, and research spotlights.


⚖️ Our promise

Plain talk, practical evidence, and transparent limitations. We disclose conflicts, avoid vendor hype, and always describe the context, dataset, and validation design behind any result.


Hosted on Acast. See acast.com/privacy for more information.

]]> false Maryna Kuzmenko info+68c5cbd124c6e67e682a0119@mg-eu.acast.com 68c5cbd124c6e67e682a0119 future-of-agriculture https://assets.pippa.io/shows/cover/1757793119454-faca0a67-ba99-433b-9109-3e6ae8300396.jpeg https://feeds.acast.com/public/shows/future-of-agriculture Future of Agriculture Podcast AI in Agriculture — Why the Fastest Wins Are Off the Field AI in Agriculture — Why the Fastest Wins Are Off the Field Wed, 01 Oct 2025 21:59:33 GMT 7:15 68dda4477be17a7f01a14b99 false https://www.youtube.com/@drmarynakuzmenko/videos 68dda4477be17a7f01a14b99 68c5cbd124c6e67e682a0119 Future of Agriculture - Episode 001 full 1 1 Episode 001 — AI in Agriculture: Why the Fastest Wins Are Off the Field

Future of Agriculture Podcast


In our kick-off episode, we flip the usual script on “AI in ag.” Drones and in-field robots get the headlines, but the quickest, cleanest ROI is happening after harvest—in packhouses, logistics desks, finance back offices, maintenance rooms, and compliance workflows. Why? Because these environments have low variability, fast feedback, and clear truth labels: weights, grades, defects, timestamps, barcodes, alarms. That’s exactly where AI learns fast and pays back quickly.


We explore five high-leverage areas:


  1. Packhouse & QA: Computer vision that counts, grades, and verifies labels—cutting giveaway, boosting Class I yield, and reducing chargebacks.
  2. Logistics & Cold Chain: Demand forecasts and sensor alerts that improve OTIF (on time, in full) and reduce last-minute scrambles.
  3. Finance & Admin: OCR + LLMs that read invoices, match POs/GRNs, and draft postings—shrinking cycle time and exceptions.
  4. Maintenance: PLC/SCADA log summarisation that spots chronic stoppage causes and enables predictive maintenance before a line goes down.
  5. Compliance & Traceability: “Ask your documents” copilots for SOPs, BRCGS/GlobalG.A.P./HACCP gap checks, and instant audit prep.


We also get philosophical about why not the field (yet): open systems, slow ground truth, and high ecological stakes. Inside the packhouse, experiments are safer and reversible; you can adjust a tolerance or a label rule and see the outcome today—not at season’s end.



Simple rollout you can copy


  • Pick one line, one SKU, one KPI (e.g., giveaway 3.2% → 2.4%).
  • Use data you already have: 12–24 months of orders, basic QC sheets, downtime codes, clean SKU/pack masters.
  • Pilot two tools: a vision check on one station + invoice OCR & doc Q&A.
  • Track weekly: Class I share, giveaway, PPH/UPH, invoice cycle time, exception rate, OTIF.
  • Keep it safe: human-in-the-loop for grade/safety, log all AI actions, EU/UK data residency, and no vendor training on your private data by default.



Key takeaways

Fields are complex; packhouses are controlled. Start where entropy is low and feedback is instant.
AI thrives on data exhaust (weights, alarms, timestamps). You already produce it—now put it to work.
Small, reliable improvements (-1% giveaway, fewer stoppages, faster postings) compound across a season.
Good, consistent data beats “perfect” data. Culture and SOPs matter more than shiny sensors.



Prompts to try tomorrow


  1. “Summarise yesterday’s Line 2 QC. Top 3 defects + quick fixes.”
  2. “Cluster chiller alarms (7 days) and suggest likely parts to check.”
  3. “Build tomorrow’s pick list from confirmed orders and flag shortages.”
  4. “Compare SOP-PACK-014 with BRCGS 4.10.2 and list gaps in plain English.”



Who this episode is for


Growers, packhouse managers, fresh-produce exporters/importers, co-ops, and agrifood SMEs tasked with throughput, quality, and compliance—on tight budgets and tighter timelines.


Support & share


If this helped, follow/subscribe and share Episode 001 with someone running a packhouse or a produce desk. Tell us your biggest off-field pain point—“label errors on 250g berries,” “invoice backlog,” “chiller alarms.” We’ll turn a few into simple 90-day plans you can run next month.

Hosted on Acast. See acast.com/privacy for more information.

]]> Episode 001 — AI in Agriculture: Why the Fastest Wins Are Off the Field

Future of Agriculture Podcast


In our kick-off episode, we flip the usual script on “AI in ag.” Drones and in-field robots get the headlines, but the quickest, cleanest ROI is happening after harvest—in packhouses, logistics desks, finance back offices, maintenance rooms, and compliance workflows. Why? Because these environments have low variability, fast feedback, and clear truth labels: weights, grades, defects, timestamps, barcodes, alarms. That’s exactly where AI learns fast and pays back quickly.


We explore five high-leverage areas:


  1. Packhouse & QA: Computer vision that counts, grades, and verifies labels—cutting giveaway, boosting Class I yield, and reducing chargebacks.
  2. Logistics & Cold Chain: Demand forecasts and sensor alerts that improve OTIF (on time, in full) and reduce last-minute scrambles.
  3. Finance & Admin: OCR + LLMs that read invoices, match POs/GRNs, and draft postings—shrinking cycle time and exceptions.
  4. Maintenance: PLC/SCADA log summarisation that spots chronic stoppage causes and enables predictive maintenance before a line goes down.
  5. Compliance & Traceability: “Ask your documents” copilots for SOPs, BRCGS/GlobalG.A.P./HACCP gap checks, and instant audit prep.


We also get philosophical about why not the field (yet): open systems, slow ground truth, and high ecological stakes. Inside the packhouse, experiments are safer and reversible; you can adjust a tolerance or a label rule and see the outcome today—not at season’s end.



Simple rollout you can copy


  • Pick one line, one SKU, one KPI (e.g., giveaway 3.2% → 2.4%).
  • Use data you already have: 12–24 months of orders, basic QC sheets, downtime codes, clean SKU/pack masters.
  • Pilot two tools: a vision check on one station + invoice OCR & doc Q&A.
  • Track weekly: Class I share, giveaway, PPH/UPH, invoice cycle time, exception rate, OTIF.
  • Keep it safe: human-in-the-loop for grade/safety, log all AI actions, EU/UK data residency, and no vendor training on your private data by default.



Key takeaways

Fields are complex; packhouses are controlled. Start where entropy is low and feedback is instant.
AI thrives on data exhaust (weights, alarms, timestamps). You already produce it—now put it to work.
Small, reliable improvements (-1% giveaway, fewer stoppages, faster postings) compound across a season.
Good, consistent data beats “perfect” data. Culture and SOPs matter more than shiny sensors.



Prompts to try tomorrow


  1. “Summarise yesterday’s Line 2 QC. Top 3 defects + quick fixes.”
  2. “Cluster chiller alarms (7 days) and suggest likely parts to check.”
  3. “Build tomorrow’s pick list from confirmed orders and flag shortages.”
  4. “Compare SOP-PACK-014 with BRCGS 4.10.2 and list gaps in plain English.”



Who this episode is for


Growers, packhouse managers, fresh-produce exporters/importers, co-ops, and agrifood SMEs tasked with throughput, quality, and compliance—on tight budgets and tighter timelines.


Support & share


If this helped, follow/subscribe and share Episode 001 with someone running a packhouse or a produce desk. Tell us your biggest off-field pain point—“label errors on 250g berries,” “invoice backlog,” “chiller alarms.” We’ll turn a few into simple 90-day plans you can run next month.

Hosted on Acast. See acast.com/privacy for more information.

]]>