Analytics to Action: ecoation’s IPM Outbreak Projection Warns Growers Weeks Before Impact

When growers bring an OKO into their greenhouse, they get more than a machine. The OKO collects a large amount of data in each row and our AI-driven platform uses this data to equip growers with powerful insights and reports. In addition to our OKO robot and AI-driven platform, the human component of our service packages includes dedicated agronomists and digital farmers to make sure our customers are getting value from the data. It’s the combination of ecoation’s product, platform, and service that enables a focused, prevention-over-cure mindset that is used to project trends and alert growers of potential issues weeks in advance.


Detection implies the problem already exists, leaving growers one step behind. IPM Outbreak Projection, on the other hand, enables growers to get ahead of issues with AI-driven projection that prompts proactive action before the issue gets out of hand.


Pest images © Biobest Group NV


Translating Analytics Into Action:

Being able to know if a pest issue is going to escalate before it reaches the economic action threshold, or even more critical, the economic injury threshold and impacts production is a struggle as old as agriculture itself. ecoation’s IPM Outbreak Projection technique aims to address this far-reaching grower pain point by providing an early warning to greenhouse decision-makers that they are likely to exceed these thresholds in the coming weeks if they do not take proper action. To make this possible, ecoation has developed a combined human-AI technique for flagging subtle pattern changes in the pest and disease population dynamics through scouting, treatment data, and micro-climate and light levels for each square meter of the greenhouse – which combine into a leading indicator of a developing pest or disease outbreak.


The next steps are tailored to our customers. We know that our growers know their greenhouses best, and our goal is to provide information that brings grower attention to when there might be cause for alarm in the near future. When ecoation provides growers with this type of early warning, it can enable earlier action and they may be able to better contain the outbreak using biological agents or use very targeted treatments in a smaller area. Bypassing spending on chemical intervention not only mitigates further crop stress but also avoids wasting the initial investment made in the current population of beneficials. This technique has demonstrated the ability to flag and warn of an emerging outbreak as much as 3 weeks in advance of the pest becoming a recognized issue in a greenhouse.


IPM Outbreak Projection in Action…

Commercial scouting data is subjective and often noisy, meaning that when viewing data we typically see jumps up and down week over week,” shares ecoation VP Data Science, Greg Stewart.These excursions can be misleading if not treated carefully. ecoation's analysis technique is able to reliably see the trends through the noise by combining multiple data sources (scout labels, micro-climate, and treatment) into a single estimate of the pest population's trend.


The chart (right) contains scouting data from a greenhouse experiencing a growing outbreak starting at week 43 and continuing through week 50. However, the apparently decreasing period between weeks 46-49 could be misleading if taken out of context. ecoation analysis was able to recognize the multi-week growing trend starting at week 44 in spite of the up and down “noise” in some of the weekly data.


Supporting the Human Eye in Greenhouse Operations:

ecoation is always looking for new, improved ways to act as an extension of our customers’ teams. Innovating at the intersection of human knowledge and machine precision means that not only do we collect massive amounts of data in each row and use our powerful AI-driven platform to compute insights, but we also have a capable and invested team in place to do weekly check-ins with our growers, review insights, gather feedback, and ultimately make sure anomalies in the data don’t go unnoticed. In the context of IPM, we can use our projection technique to keep our Customer Success team up-to-date on what’s happening on the ground, ready to send warning notifications at the earliest signs of a heightened pest and/or disease risk. This way, ecoation is able to provide timely alerts to give growers the information they need to make their decision and take action.


Putting It All Together:

The volume of data generated from greenhouse operations is astronomical. At ecoation, we continue to build out tools that capture what's happening post-by-post with SqM precision granularity and use AI to derive insights that save our customers time, money, and stress. In this next step with IPM Outbreak Projection, we’re supplying these insights in simple ways that support timely decisions. To deliver on this, we have our new technique running across greenhouses all over the world and processes in place to get growers the right numbers, at the right time, so they can make the best decision.