Automations, be it rules or more complex machine learning algorithms, are usually the end goal of your IOT solution. Without this aspect, it’s hard to really speak about an IOT solution.
Automations can be simple: getting a notification when the battery needs to be replaced or when a threshold has been reached. They can also be much more complex, requiring sensor fusion (combining multiple sensor readings) and the use of some sort of smart algorithm able to learn by itself.
Many cloud platforms already support some level of automation:
- Rule for AWS IoT provides both simple condition based rules and machine learning algorithms
- Azure provides an advanced machine learning infrastructure but does not provide any out-of-the-box conditional solutions.
- Google provides a very good machine learning library called ‘TensorFlow‘, a set of tools for analytics and a extensive cloud solution.
- IBM watson is great at both basic conditional based as well as highly advanced machine learning.
What we can do
No matter which platform, the leaning curve is usually pretty high and it’s not always clear where to start or how to proceed. At Elastetic, we are looking forward to help you out. Let us know your needs and we can:
- Do an analysis of your needs and make a proposition for an optimal solution tailored to your needs.
- Implement the resulting automations in the platform of your choice.
- Build custom automation solutions, according to your specific needs.