Grid-EYE Application vertical
Application vertical - Temperature gradient for position tracking and counting
Temperature gradient data from Grid-EYE can be used detect the presence and track the location of humans in a given field of view.
Infrared sensors like PIR sensor and Grid-EYE are optimized to detect IR around the wavelength ~10µm - which corresponds to wavelength of thermal radiation emitted by human body. In smart building use cases, such sensors are therefore often used for applications like automated lighting actuation, intrusion detection, etc. Compared to standard PIR sensor, Grid-EYE can provide intelligent insights on proper utilization of space.
In modern smart building infrastructure, there can be hundreds of sensors deployed for people detection and counting. Such a deployment can have many pitfalls e.g. Congested M2M data pipelines, AI/ML computation resources, Data privacy concerns. The 64-pixel resolution and a peak image refresh rate of 10 Hz provide an optimal compromise to the above cited problems.
Privacy: No private data – identity or gender – can be deciphered.
Data footprint: Grid-EYE generates less than 1kb data per second, bringing down data transportation cost and network congestion compared to higher resolution devices.
Computational resource: Detection and tracking can be run locally on constrained devices like microcontrollers.
New computational techniques leveraging Machine Learning algorithms can also be used on Grid-EYE to lead to smart people counting and tracking solutions.
For more information, consult GridAI evaluation kit.
