April 7, 2025
•
5 min
read
Phenotyping lies at the heart of agricultural innovation. The accurate and regular characterisation of key plant traits throughout the growth cycle is essential to the success of any experiment.
Effective phenotyping requires the collection of large amounts of data. This is due to several factors:
• Multiple traits of interest (observation parameters) typically need to be monitored. These depend on the crop species and the specific goals of the trial.
• Traits are recorded at different growth stages, sometimes even daily.
• Measurements can be taken on different levels at a field level for on-farm trials, at a plot level or at a plant/fruit level to detect even more advanced traits.
• For statistical robustness, subplots must be replicated.
This extensive data collection can be highly time-consuming for trial teams.
In recent years, new "high-throughput" phenotyping techniques have emerged. These approaches, which leverage modern technologies and IT, are capable of producing large volumes of phenomic data either automatically or with minimal human intervention.
The purpose of this article is to present some of these phenotyping methods, their benefits, and their applications in agricultural research and innovation.
Contents:
I. Main Methods of High-Throughput Phenotyping
2. Connected Sensors and Automated Platforms
II. Managing the Large Data Volumes Generated by High-Throughput Phenotyping
Conclusion: High-Throughput Phenotyping – A Key Asset to Address Tomorrow’s Challenges
Imaging is often the first thing that comes to mind when discussing high-throughput phenotyping. Over the past 15 years, imaging technologies have flourished in agriculture, with a wide range of applications.
Various types of sensors are used to capture these images, including:
In terms of deployment, two main modes of sensing exist:
Proximal sensing typically provides higher-resolution images (due to proximity) and allows varied viewing angles. It's well-suited for vegetable or fruit crops as it can generate plant-by-plant images. It’s also relatively easy to implement, as it doesn’t require drones.
Remote sensing enables phenotyping over entire experimental fields, clearly identifying differences between subplots with different treatment combinations. Compared to proximal sensing, it allows large-scale, rapid coverage—ideal for arable crops like cereals, maize, and oilseeds.
The images collected by sensors are analyzed through dedicated software pipelines to generate phenotyping maps and extract trait data. Examples of outputs include:
Several providers propose solutions based on different technologies, different possibilities to scale up at large level (when the company has hundreds of thousands of plots to phenotype), based on different sensors or hardware. In our industry we often deal with Hiphen, Alteia, Eiwa, Vito and Phenospex.
AI-powered models are used to automatically interpret sensor data and extract phenotypic traits. Read more on AI for crop innovation.
They can help you establish good methodology to capture data, even tailored hardware designed for your specific traits. Integration of the high-throughput phenotyping platform with your central research database is key. When the agronomy testing software connects seamlessly, it can send the trial protocol and environment information to the platform, and receive back the trait data to evaluate your material.
The development of IoT over the past 20 years has brought connected sensors into many sectors, including agriculture.
Connected sensors allow real-time phenotyping without the need for manual intervention. This represents a major time-saving opportunity for data collection teams.
Some companies have gone even further, offering fully automated phenotyping platforms.
On-field systems can collect data in real time from the environment (rainfall, temperatures, soil reflectance…) or in the laboratory.
They can control specific environmental conditions (e.g. soil water availability, fertilisation levels) and monitor parameters in real time, such as:
This enables researchers to study how genotypes respond to different environmental conditions (GxE interactions). For instance, one could compare several varieties under water stress.
Several researchers have shared how valuable this type of system is in monitoring their trials. For example, at CTIFL, the French Interprofessional Technical Centre for Fruits and Vegetables, they integrate phenomics from the analytical lab:
“To analyze fruit taste quality, several measurements are taken in the lab using an automated unit: sugar content, acidity level, and flesh firmness.
The operator enters experiment and material identifiers into the unit, which enables direct re-import of results into the database.
By streamlining the link between measurement tools and the central database, Doriane’s software simplifies data entry, reduces the risk of errors, and saves valuable time!”
Read the full case study: CTIFL fruit and vegetable testing
As we've seen, high-throughput phenotyping produces vast amounts of data, sometimes in real time. Furthermore, multiple phenotyping methods are often used simultaneously within a single trial to monitor various traits.
Before they can begin data analysis, researchers must:
It’s also vital that the data can be archived (e.g. for future audits).
Managing all this in Excel can be extremely tedious and time-consuming—especially with large datasets.
That’s why it’s essential to use dedicated software tools like Bloomeo agronomy testing software, designed specifically for managing agricultural trials. In Bloomeo, experimental data is structured, centralized, secured, and shared collaboratively. Data integration is supported through several channels:
This means data analysis can begin much faster, with minimal risk of error and minimal effort spent validating the dataset.
High-throughput phenotyping opens up new possibilities in agricultural experimentation by dramatically reducing the time teams need to spend in the field to collect large volumes of data. In practical terms:
• Significant labour savings are possible.
• Researchers can focus more on analysis rather than data collection.
• Trial designs can be expanded to include more replications or treatment combinations, resulting in more robust results.
However, these techniques also have their limitations:
• Initial investments and operating costs can be high.
• Not all crop types or trial formats (e.g. greenhouse vs. open-field) are compatible.
• Not all traits can be phenotyped using automated techniques. Manual measurements are still needed for certain traits (e.g. fruit weight in apple trees).
At present, high-throughput phenotyping doesn’t fully replace traditional methods, but it complements them—and its role in crop innovation will only continue to grow.
Finally, as we’ve seen, high-throughput phenotyping must be integrated into a robust data management system to fully unlock its potential and deliver reliable, actionable insights.
Jean-Baptiste L.
Agronomist
Did you find this article useful? Share it!
Webinar Replay
Next webinar
April 24, 2025
Unlock the power of multi-factorial trials! Learn from our agronomist and IT expert to design, test, and analyze interaction effects for smarter crop research.
Patricija Levickaite
IT expert & Agronomy engineer
April 7, 2025
•
5 min
read
Phenotyping lies at the heart of agricultural innovation. The accurate and regular characterisation of key plant traits throughout the growth cycle is essential to the success of any experiment.
Effective phenotyping requires the collection of large amounts of data. This is due to several factors:
• Multiple traits of interest (observation parameters) typically need to be monitored. These depend on the crop species and the specific goals of the trial.
• Traits are recorded at different growth stages, sometimes even daily.
• Measurements can be taken on different levels at a field level for on-farm trials, at a plot level or at a plant/fruit level to detect even more advanced traits.
• For statistical robustness, subplots must be replicated.
This extensive data collection can be highly time-consuming for trial teams.
In recent years, new "high-throughput" phenotyping techniques have emerged. These approaches, which leverage modern technologies and IT, are capable of producing large volumes of phenomic data either automatically or with minimal human intervention.
The purpose of this article is to present some of these phenotyping methods, their benefits, and their applications in agricultural research and innovation.
Contents:
I. Main Methods of High-Throughput Phenotyping
2. Connected Sensors and Automated Platforms
II. Managing the Large Data Volumes Generated by High-Throughput Phenotyping
Conclusion: High-Throughput Phenotyping – A Key Asset to Address Tomorrow’s Challenges
Imaging is often the first thing that comes to mind when discussing high-throughput phenotyping. Over the past 15 years, imaging technologies have flourished in agriculture, with a wide range of applications.
Various types of sensors are used to capture these images, including:
In terms of deployment, two main modes of sensing exist:
Proximal sensing typically provides higher-resolution images (due to proximity) and allows varied viewing angles. It's well-suited for vegetable or fruit crops as it can generate plant-by-plant images. It’s also relatively easy to implement, as it doesn’t require drones.
Remote sensing enables phenotyping over entire experimental fields, clearly identifying differences between subplots with different treatment combinations. Compared to proximal sensing, it allows large-scale, rapid coverage—ideal for arable crops like cereals, maize, and oilseeds.
The images collected by sensors are analyzed through dedicated software pipelines to generate phenotyping maps and extract trait data. Examples of outputs include:
Several providers propose solutions based on different technologies, different possibilities to scale up at large level (when the company has hundreds of thousands of plots to phenotype), based on different sensors or hardware. In our industry we often deal with Hiphen, Alteia, Eiwa, Vito and Phenospex.
AI-powered models are used to automatically interpret sensor data and extract phenotypic traits. Read more on AI for crop innovation.
They can help you establish good methodology to capture data, even tailored hardware designed for your specific traits. Integration of the high-throughput phenotyping platform with your central research database is key. When the agronomy testing software connects seamlessly, it can send the trial protocol and environment information to the platform, and receive back the trait data to evaluate your material.
The development of IoT over the past 20 years has brought connected sensors into many sectors, including agriculture.
Connected sensors allow real-time phenotyping without the need for manual intervention. This represents a major time-saving opportunity for data collection teams.
Some companies have gone even further, offering fully automated phenotyping platforms.
On-field systems can collect data in real time from the environment (rainfall, temperatures, soil reflectance…) or in the laboratory.
They can control specific environmental conditions (e.g. soil water availability, fertilisation levels) and monitor parameters in real time, such as:
This enables researchers to study how genotypes respond to different environmental conditions (GxE interactions). For instance, one could compare several varieties under water stress.
Several researchers have shared how valuable this type of system is in monitoring their trials. For example, at CTIFL, the French Interprofessional Technical Centre for Fruits and Vegetables, they integrate phenomics from the analytical lab:
“To analyze fruit taste quality, several measurements are taken in the lab using an automated unit: sugar content, acidity level, and flesh firmness.
The operator enters experiment and material identifiers into the unit, which enables direct re-import of results into the database.
By streamlining the link between measurement tools and the central database, Doriane’s software simplifies data entry, reduces the risk of errors, and saves valuable time!”
Read the full case study: CTIFL fruit and vegetable testing
As we've seen, high-throughput phenotyping produces vast amounts of data, sometimes in real time. Furthermore, multiple phenotyping methods are often used simultaneously within a single trial to monitor various traits.
Before they can begin data analysis, researchers must:
It’s also vital that the data can be archived (e.g. for future audits).
Managing all this in Excel can be extremely tedious and time-consuming—especially with large datasets.
That’s why it’s essential to use dedicated software tools like Bloomeo agronomy testing software, designed specifically for managing agricultural trials. In Bloomeo, experimental data is structured, centralized, secured, and shared collaboratively. Data integration is supported through several channels:
This means data analysis can begin much faster, with minimal risk of error and minimal effort spent validating the dataset.
High-throughput phenotyping opens up new possibilities in agricultural experimentation by dramatically reducing the time teams need to spend in the field to collect large volumes of data. In practical terms:
• Significant labour savings are possible.
• Researchers can focus more on analysis rather than data collection.
• Trial designs can be expanded to include more replications or treatment combinations, resulting in more robust results.
However, these techniques also have their limitations:
• Initial investments and operating costs can be high.
• Not all crop types or trial formats (e.g. greenhouse vs. open-field) are compatible.
• Not all traits can be phenotyped using automated techniques. Manual measurements are still needed for certain traits (e.g. fruit weight in apple trees).
At present, high-throughput phenotyping doesn’t fully replace traditional methods, but it complements them—and its role in crop innovation will only continue to grow.
Finally, as we’ve seen, high-throughput phenotyping must be integrated into a robust data management system to fully unlock its potential and deliver reliable, actionable insights.
Jean-Baptiste L.
Agronomist
Did you find this article useful? Share it!
