Seeking European partners to develop detection application(S) for greenhouse sector

As a participant of the upcoming symposium “2nd GCCIR Matchmaking Symposium” organized by Innoget and GCCIR next November 25th, 2019 in Barcelona (Spain), the company Stream is seeking European partners to meet at the Barcelona event to develop a joint collaboration project.

Stream is specifically interested in European partners to develop detection application(s) for greenhouse sector. Further details are described below.

In case you are interested in meeting Stream in Barcelona to discuss a potential collaboration, please register for free to the symposium clicking the “Register for the event” button above.

Attachments

Details of the Technology Call

Stream is an analytics company that specializes in machine learning and spectroscopy to change the way detection is done in the world. Stream value is to provide highly accurate and timely machine learning predictions for industry. The company’s data science team has developed advanced neural nets designed to leverage both spatial and spectral data. Stream also provides tools that enable non-data scientists to build prediction models, without the need of knowledge or experience with machine learning.

Models that are developed are automatically hosted for distribution and use. In addition, Stream provides a marketplace to support instant commercialization of new models world-wide.

The company analyzes images from cell phones to scans from spectrometers. Detect the presence of a disease, fungus, or predict a specific value like the percentage of protein in barley, nutrients or pesticide residue, as possible examples. This applies to quality control in food processing or any other number of targets. This eliminates the need to wait for test results coming back from the lab.

The company’s platform enables non-data scientists with samples to build their own custom analytics models, simply by taking pictures and pushing the train button. Stream is specialized in allowing developers, with limited knowledge or experience in Machine Learning to automatically build Machine Learning models from images and scanned data. The deep learning neutral nets have been optimized to return highly accurate predictions from this type of data. A simple API call returns the prediction values, to be embedded directly into an application. The version 1.0 of the analytics platform has been released. There are limited algorithms or models that have been build, as the platform is new.

Stream is looking to expand its reach and use of its analytics engine into the greenhouse sector and are specifically looking for partners that could collaborate to develop detection application(s) for this sector. This could include various types of crops, and Stream is looking for companies that could help with the ability to get cameras moved efficiently around a facility, so as to be able to image an entire crop canopy.

The company is looking to partner with companies who provide the ground truth data for various disease or other targets that are valuable through ought Europe. The partners Stream is looking for:

• Have a relationship with end user customers or their business model is such that they provide solutions to this market. Once Stream and the partner work together to develop a series of algorithms, Stream’s business model is to provide ongoing predictions to the reseller or developer (the Partner) and or other resellers.
• Ideal partners might also be able to provide or source 1) those who can effectively provide training data, e.g. hundreds of images of disease plants, for the analytics engine, 2) those that can provide a novel way to move imaging systems (cameras) throughout a greenhouse.

Preferences / Requirements for Potential European Partner(s)

The following are the important aspects Stream is looking for in a European partner:

• Stream is looking for experts that would know the type of crops that are of value, the disease or targets of value and the way that ground truth data is collected. They would be able to determine if the new algorithms we jointly developed would be able to outperform the current method used in the industry.
• It would be even better if they had access to parts of the market that would be willing to test the company’s new system.
• Stream is interested in knowing if the soil is used in greenhouses in Europe and if so if there are companies that have experience in soil management and soil detection – access to a wet chemistry lab would be fantastic.
• Stream is also interested in companies that are doing the same type of activities but NOT in a greenhouse – precision agriculture.