The Magic of Phone QC

We felt like magicians when we demoed our wheat assessment technology for the first time, QCing more than 1000 grains of wheat with a phone in a minute.

At Evoke (Austral Asia's primer Agtech event ), we experienced first-hand the excitement, interest and enthusiasm that AI assessment can generate.

Seeing is Believing

We are certainly not the first to attempt assessing grains with AI. Millions have been spent over a decade now to build imaging and AI solutions for evaluating grain. Yet, we do not see any in operation. Grain is laborious, and hand inspected all over the world.

There are quite a few attempts all over the world to crack the problem of grain assessment, notably scanner-based solutions from Zoom Agri (Argentina), AgNext, Intello Labs, Nebula (India), Camera-based solutions from Cropify (Australia) and Machine AI Solutions from Indyn (Australia). There are also attempts by major grain corporations to build their in-house solutions. 2024 is set to be to a landmark year as many of these solutions are undergoing trials and are at various stages of development. It is undoubtedly an exciting time for AI assessment.

We are betting on phones

Like most teenagers - we believe that smartphones are all-powerful. Thier cameras are fantastic, they connect to the internet, and all the clever stuff happens there. There is no need for expensive cameras and conveyer belts, and hoppers to arrange the grain in rows, so they do not touch each other. We do not need to do that. We can throw them together - a good thousand or more grains and one phone photo will do the job. Cloud-based AI can do the rest.

There is no need to arrange or align

Current attempts at AI assessment require grains not to touch each other. The grains are either arranged or aligned so that the grains can be individually cropped and identified. GoMicor's AI imaging technology can assess grains that are touching, and this makes a significant difference.

More importantly, our patent-pending technology makes it possible to create ideal imaging conditions for AI to learn fast. It makes it possible to train high accurcy AI apps with hundreds of images instead of the thousands required by the competition. AI applications that took years to build can now be built in weeks.

One of the challenges in AI assessment is not the assessment of common grain defects but the identification of inclusions such as weed seeds - that differ from state to state. These samples are hard to collate, and grain corporations see their collection of these treasure chests, particularly the thousands of images they have so laboriously collected over the years and their expertise in differentiating defects. These investments are likely to lock them into solutions they have invested in.

DIY AI Apps

The combination of advanced imaging technologies that we have developed and new AI methods makes it possible to build AI apps rapidly with a small sample set.

The next decade will revolve around Data-Centric Artificial Intelligence, predicts AI pioneer Andrew Ng. We may not need million of noisy samples if we have 50 well-crafted ones.

GoMicro's patent-pending imaging technology can create well-crafted data sets. This is our secret sauce. Our superior imaging technology makes it possible to build rapidly, test rapidly, and deploy AI assessment applications in the hundreds - each tailored to specific grain types, seasonal variations, and buyer requirements.

GoMicro AI Apps can now assess wheat, corn, lentils and soy; the list is growing fast. GoMicro is in the business of connecting the dots. This can only be done if payers at all points of the grain chain can QC to the same standards. Phone based QC solutions will make it possible for farmers worldwide to QC accurately in a way that is faster and better but cheaper than manual QC. Trials are well underway in India, Ghana and Indonesia.

Australian farmers can now QC on farm

Most farmers have phones. They can now QC their grain on-farm with a low-cost GoMicro device; before they truck it all the way to the receiving stations. We can help them decide where they would get the best price for their grain and if thier grain will meet the standards required at recyclable stations. If they don't, they can decide to clean it with the help of GoMicro AI Assessment.

Joe had a couple of lentil deliveries rejected

QC is about the reduction of risk

Many may not see it that way - particularly if they are schooled in the principles of precision. But that is what QC is all about. Farmers and grain buyers have a good sense of quality. Many can guess the grain by just looking at it. Yet, if you are buying and selling grain by the tons - visual assessment is too risky - that risk must be reduced at both ends.

Rapidly evolving grain trading practices

While major grain companies, with investments in the infrastructure of transport, shipping and trading, manage most of the grain trade - we notice the emergence of nimbler players - enabling the digital trading of grain. Clear Grain, GrainStre, Rayglen, Igrain, GrainInex, Mercaris,CHS,CombyneAg, AgriDigital and many more. Also, many farmers now store and trade thier grains at better prices using online platforms. A severe constraint in the growth of digital grain trading is the uncertainty around grain quality.

If grain quality can be assessed on-farm and all along the chain - using the same instrument, standard and technology - the game of grain trade will transition to the next level, connecting buyers and sellers globally with new levels of efficacity

In this context, companies and farmers who embrace advanced QC technologies will enjoy significant advantages - over those who are too busy - assessing them one by one by eye.

________________________________________________________________

Dr.Sivam Krish CEO/GoMicro Ltd.