ImagoAI Real-Time On-site Soybean Testing

The _

Customer is one of the world’s largest soybean processors. ImagoAI is the pioneer of AI-driven onsite real-time food quality analysis. ImagoAI is backed by Google for startups & Techstars and in a short span of time has onboarded some of the world’s largest food processors on its AI platform.

‍

.

‍

Losses to customer due to inadequate soybean quality analysis.

Below are the key losses faced before ImagoAI:
● Damaged soybean produces oil with high FFA (Free Fatty Acids) which has a lower market value.
● ~1% discount is levied on high FFA oil leading to millions of dollars in lost revenue.
● Storing damaged soybean with good quality soybean leads to a loss in oil yield.
● Missing the opportunity to discount/re-route poor-quality trucks.

‍

‍

Challenges faced by Customer in soybean quality measurement.

● Human bias in measuring Green Damage, Arjidu and Fermented Damage
● Inaccurate: Human bias leads to inaccurate and unreliable results.
● Destructive: Soybean kernels need to be cut into halves to look for damages.
● Time consuming: Time taken per sample for damage analysis is 10-15 minutes

Why Choose ImagoAI
ImagoAI is the pioneer of AI-driven onsite real-time food quality analysis. ImagoAI is backed by Google for startups & Techstars and in a short span of time has boarded some of the world’s largest food processors on its AI platform. ImagoAI is the only solution in the market that measures all the soybean quality parameters in an accurate, consistent, real-time, & non-destructive manner.

How Customer can leverage ImagoAI’s solution
● Damaged soybean can be stored in separate bins to prevent them from infecting good-quality grains.
● Damaged soybean when mixed in the right proportion with good quality soybean can produce optimum quality oil.
● Trucks containing damaged soybeans can be rerouted to competitors or rejected.
● Right discounting can be applied to vendors.
● Ultimately, millions of dollars saved per facility by producing high-quality soybean oil

‍

The Results
Below are the graphs illustrating the high accuracy and robustness of our solution on Soybean 6 quality parameters:

Moisture

● Current Method : Oven based or NIR
● Current Time Taken per Sample : ~ 1 min

<insert graph image>

Oil

● Current Method: NIR
● Current Time Taken per Sample: ~ 1 min

<insert graph image>

Protein

● Current Method : NIR
● Current Time Taken per Sample : ~ 1 min

<insert graph image>

‍

Green Damage
● Current Method: Human Visual
● Current Time Taken per Sample: ~ 10 - 15 mins
● Issues: Destructive, Time Consuming, Inaccurate, Human Biased

<insert graph image>

Fermented Damage
● Current Method: Human Visual
● Current Time Taken per Sample: ~ 10 - 15 mins
● Issues: Destructive, Time Consuming, Inaccurate, Human Biased

<insert graph image>

Arjidu
● Current Method: Human Visual
● Current Time Taken per Sample: ~ 10 - 15 mins
● Issues: Destructive, Time Consuming, Inaccurate, Human Bias

<insert graph image>

‍

‍

‍

‍