Is this technology a fit?

If consistency or quality is important in your business, then perhaps a reasonably priced detection system that could detect “targets” you cared about would be welcomed.

If knowledge about the state of your product can help increase yield or decrease input costs, this technology could be helpful.

If you provide maps or advice to the agriculture industry, or you want a better soil testing system or advanced analytics from airborne data – we should talk.

If you have data from existing equipment that could be used to train a machine learning model, you could leverage that data into real revenue.

There are four broad categories of stakeholders that can benefit as spectroscopy and machine learning bring detection Aapps to everyday business.

Business Owners/Quality Control Managers

Business owners concerned with quality control, or who just want to know what is going on at different stages of their product lifecycle, can enjoy faster results and lower costs. Our customers who have used their own samples to build Aapps with soils, seeds, or plants, together with agronomists advising on crop related decisions, will find this useful. Anyone operating a greenhouse or producing a product through multiple steps could likely also benefit from instant detection results.

(Read more...)

For example:

Finding corn blight might be useful for a farmer in Iowa.

Finding protein in wheat could be applicable to a grain farmer in western Canada.

Camera, Spectrometer and Digital-Signal-Device Manufacturers 

Camera manufacturers can benefit by having images from their cameras for scientific applications. RGB cameras can be used to capture useful training data and once a new Aapp is built and shared through the online store, the value of the hardware continues to increase.

Spectrometer manufacturers benefit by having their entire range of devices be capable of providing the data to both train and test across a broad range of applications. Spectrometers will find many new uses, outside the scope of just labs and research.

(Read more...)

In other words, the value of a camera or spectrometer as a scientific instrument increases. The whole idea is to increase the value of the hardware in the hands of customers.

  • Multispectral and Hyperspectral camera manufacturers.
  • Spectrometer manufacturers – add your lineup of hardware to the supported list.
  • Microscope manufacturers.
  • Ultra-sound and other signal processing devices.

Selling your hardware:

The main reason to integrate your lineup of cameras and spectrometers is to sell more units.  It also shows your customer base you are on the leading edge of technology affecting your market.

There are no hardware changes needed; check out our published, open source API.

Work with Stream’s marketing team to offer bundled Aapp developer discounts to your marketing campaigns.

Streams support for the Open Network Spectral Interface (ONSI) makes it easy for hardware manufacturers to support a single open source API, allowing software developers to communicate effectively with supported hardware.

Researchers, University and Colleges Students

While there are already fantastic tools that work with spectral data, Stream is offering another tool designed to save you a ton of time when creating machine learning models. You only need labelled training data, and Stream’s neural nets will process the data and build a detection algorithm, or Aapp, without the need to know about how to build a algorithm at all. And it can all be free!

(Read more...)

If you have existing spectrometers of any sort or are working with data from any part of the spectrum, perhaps the use of machine learning and neural nets to quickly create a detection algorithm would be extremely helpful.
If you have or can produce labelled training data from your spectrometer or similar type of device, you can upload your files into the Stream Analytics Engine to build a model, or what we call an Algorithm application (Aapp). Once an Aapp is built, it can be used over and over for fast predictions. Build regression or multi-class Aapps just by uploading your spectral data and pushing the “Train” button – for free.

If you have existing files – you’re just a couple of hours away from a new Aapp.

If you are using an ‘ONSI’ compliant spectrometer – even the upload is automated.

While you might have your own reason to use machine learning to build an Aapp – here is one more. Stream pays a royalty to anyone who creates an Aapp and sells it through our online Aapp store.

As a student, if you have access to a spectrometer or camera, along with samples, building an Aapp and collecting royalties from your work might be of interest.

If you have access to sample data or know how to take spectral scans or pictures with a camera, you can create an Aapp. As other people subscribe to your Aapp, Stream will pay you a royalty as part of our Creator Program.

This is why students and staff from colleges and universities are setting up their free accounts.

Value Added Resellers / System Integrators 

You already know how to install and service various cameras, lighting or other hardware. Now allowing customers to build their own prediction Aapps could provide new opportunities.

If you sell cameras for machine vision or spectrometers, our analytics engine allows anyone who can operate your data capture device to build a world-class detection algorithm. This has the potential to turn your camera into a highly effective scientific instrument.

(Read more...)

If you are in the aerial photography business – we want to talk about a specialized wide-swath multispectral camera.

If you provide maps or advice to the agricultural industry, or you want a better soil testing system, or advanced analytics from airborne data – we should talk.

Anyone who has training data from existing equipment can feed that data into our analytics engine, and once an Aapp is created, you set the price of predictions and enjoy a royalty from the Aapp’s subscription revenues.

If you work with ultra-sound or gamma detection devices or have instruments that gather signals from various parts of the spectrum – let’s talk about what you might want to auto detect. Stream can help to leverage your data beyond what the device can currently do.