Technologies Used
JavaScriptPython
Machine Learning
Sensors
Azure Machine Learning
Azure
Arduino
Vue.js
Project Team
About
Analyzing datasets to predict and prevent failures in urban property facilities. We use past events to make predictions and alert property managers on what actions to take.
Comments
Jason Chiu - 2048d
addressing a real pain point of preventive maintenance... love the demo on two sensors... it would be great to do a false alarm demo and load another set of training data to show how the learning avoid false alarm, if time permits
Krzysztof Machowski - 2048d
Thank you for your comment! It is definitely something we wanted to do, and with the ML API we were using it should be reasonably easy to integrate in a future version.
Krzysztof Machowski - 2048d
Thank you, the team from SHK were very helpful answering our questions during our intitial research.
Krzysztof Machowski - 2048d
Indeed, with the rise of low cost hardware we will likely be seeing more of these sort of smart/ low power sensors.
Dr Toa Charm - 2048d
How does it differentiate from the similar technologies in the market?
Good to have a specific problem to solve by your innovation and technology
Good to have physical objects to demo
Good visualization interface
Good use of preventive maintenance
Jainam Mehta - 2048d
Thank you for your kind comments. There is definitely movement in this area, as evidenced by many articles stating such.
Jainam Mehta - 2048d
We believe our differentiator would be our use of commodity hardware allowing for low cost sensors, as well as self-reinforcing machine learning that would reduce the work load in the future for on site staff.
Jim - 31d