From Measurement to Insight: How AI and FoodSense Generation 4 Are Transforming Chilli Quality Analysis
The food industry is generating more data than ever before, but data alone does not create value. The real challenge is turning measurements into meaningful insights that help producers, breeders, and manufacturers make better decisions. This is exactly where the combination of FoodSense Generation 4, cloud connectivity, and artificial intelligence can make a significant difference.
In a recent live demonstration, we explored how FoodSense Generation 4 enables rapid capsaicin analysis across fresh chillies and chilli-based products, while using AI-powered analysis to help users interpret and understand their results.
Why Instant Measurement Matters
One of the most common challenges in food quality testing is reliance on outdated laboratory reports and certificates of analysis. It is not unusual to encounter analytical reports that are several years old, despite significant changes that may have occurred in the product during storage.
FoodSense Generation 4 provides results in minutes, allowing users to make decisions based on the actual sample in front of them rather than historic data. Every measurement is automatically uploaded to the cloud, providing traceability, data storage, and easy access for future analysis.
The FoodSense Generation 4 Workflow
The FoodSense ecosystem consists of three key components:
FoodSense Generation 4 analyser
Mobile application with Bluetooth connectivity
Julie cloud platform for data storage and traceability
Measurements taken on the analyser are instantly transferred to the app and uploaded to the cloud, where reports can be stored, reviewed, and exported.
The addition of AI-driven analysis provides a further layer of interpretation, helping users understand what their measurements may actually mean in a practical context.
Demonstration 1: Fresh Red Bird's Eye Chilli
A fresh red bird's eye chilli sample was prepared by:
Weighing approximately 0.1 g of chilli
Extracting into 0.9 mL of buffer
Allowing extraction to occur over 24 hours
Applying 50 µL of extract to a FoodSense sensor
The measured result was:
5,570 Scoville Heat Units (SHU)
At first glance, this may appear lower than many published values found online for bird's eye chillies, which are often quoted between 50,000 and 100,000 SHU.
However, this is where AI-assisted interpretation becomes valuable.
Understanding Moisture Content
Many published Scoville values are based on dried chillies rather than fresh chillies. Since fresh chillies can contain approximately 90% moisture, direct comparisons can be misleading.
When the AI analysis accounted for the moisture content of the fresh chilli, it estimated a dry-weight equivalent of approximately:
55,700 SHU
This adjusted value falls comfortably within the expected range for bird's eye chillies and demonstrates how AI can bridge the gap between raw measurements and practical understanding.
The Importance of Domain Knowledge
Although AI is a powerful tool, it is important to recognise that large language models are not infallible.
During preparation for the demonstration, examples were identified where online AI systems:
Confused fresh chilli and dried chilli Scoville ratings
Repeated common misconceptions regarding capsaicin distribution within chilli seeds
These examples highlight the importance of combining AI with scientific expertise and validated measurement data.
AI should be viewed as an analytical assistant rather than a replacement for domain knowledge.
Sample Representativeness: Another Role for AI
A second question posed to the AI concerned sample size.
Only 0.1 g was analysed from a 25 g packet of chillies.
The AI correctly calculated that this represented approximately:
0.4% of the total sample mass
This led to an important discussion:
For homogeneous products such as blended hot sauces, a small sample may be highly representative.
For heterogeneous materials such as whole chillies, variability between individual fruits can be significant.
In this case, AI provided useful guidance by suggesting that additional samples and replicate measurements could improve confidence in the overall batch characterisation.
Demonstration 2: Green Bird's Eye Chilli
A second analysis was performed using a green bird's eye chilli from the same packet.
The measured result was:
3,380 SHU
Again, AI was used to adjust for the estimated moisture content, calculating a dry-weight equivalent of approximately:
38,000 SHU
The difference between the red and green chillies illustrates a critical point:
Not every chilli within a batch will produce identical results.
Factors such as ripeness, growing conditions, and natural biological variation can all influence capsaicin content.
Having access to rapid testing allows these differences to be measured rather than assumed.
Demonstration 3: Comparing Three Tabasco Sauce Bottles
The final demonstration involved analysing three different bottles of Tabasco sauce.
Results obtained were:
Sample | Measured SHU |
Tabasco A | 2,300 |
Tabasco B | 5,550 |
Tabasco C | 1,540 |
All results were automatically uploaded to the Julie cloud platform.
Using the comparison tools within the platform, reports from multiple samples were combined into a single comparative analysis.
Using AI to Interpret Product Differences
Rather than simply reviewing three numbers, the combined report was uploaded to an AI agent for interpretation.
The AI identified:
Significant differences between samples
Tabasco B as the hottest bottle tested
Tabasco C as the mildest bottle tested
Importantly, Tabasco C was also the oldest bottle available.
The AI suggested several possible reasons for the lower result, including:
Natural batch-to-batch variation
Differences in raw materials
Storage conditions
Capsaicin degradation over time
These observations aligned with physical observations made during the demonstration, where the oldest bottle also appeared noticeably different in colour.
Beyond Testing: Turning Data into Understanding
One of the most exciting aspects of combining FoodSense Generation 4 with AI is the ability to move beyond measurement alone.
Traditionally, food quality analysis often stops once the numerical result is obtained. However, many users still face questions such as:
Is this value reasonable?
How does it compare with industry expectations?
Is my sample representative?
Why are two samples different?
Is storage affecting product quality?
AI can help answer these questions by providing context, calculations, comparisons, and scientific reasoning.
Cloud Storage and Traceability
Every measurement performed with FoodSense Generation 4 is automatically stored in the Julie cloud platform.
This provides:
Centralised data storage
Complete traceability
Easy report generation
PDF export functionality
Historical trend analysis
Product and batch comparisons
The result is a digital workflow that supports both quality control and quality assurance activities.
Real-World Accuracy
Questions about analytical accuracy are common.
When analysing real-world food products, challenges often arise from sample heterogeneity rather than instrument performance alone.
Products such as chillies and sauces can vary naturally due to:
Agricultural variation
Processing differences
Particle distribution
Storage conditions
For this reason, repeatability within approximately 20% relative standard deviation can represent strong performance for complex food matrices.
More importantly, rapid, affordable, and repeatable testing enables users to perform multiple measurements and build statistically meaningful datasets rather than relying on a single laboratory result.
Conclusion
FoodSense Generation 4 demonstrates how modern food analysis is evolving beyond measurement and into insight generation.
By combining:
Rapid capsaicin measurements
Cloud-based traceability
Automated reporting
AI-powered interpretation
users can gain a deeper understanding of their products in real time.
Whether analysing fresh chillies, breeding programmes, hot sauces, or processed products, the combination of FoodSense Generation 4 and AI provides a practical pathway from raw data to informed decision-making.
As the food industry continues to embrace digital transformation, the ability to measure, store, analyse, and interpret data within a single workflow will become increasingly important. FoodSense Generation 4 is designed to help make that future available today.
Comments