RoadTest by Invitation: TIDA - 01069 - Review

Table of contents

RoadTest: TIDA - 01069 Adv. Motion Detector with PIR Sensors

Author: DAB

Creation date: 14 Sep 2017

Evaluation Type: Evaluation Boards

Did you receive all parts the manufacturer stated would be included in the package?: True

What other parts do you consider comparable to this product?: There was also a single PIR sensor board.

What were the biggest problems encountered?: The Demo software did not compile as advertised and the Energia Development platform was inadequate to debug the code.

Detailed Review:

I would like to thank Element 14 and Texas Instruments for allowing me to road test this board.

I really liked having the MSP432 board come with the sensor board.  Given the data monitoring needs, you really do need that level of capability.

First, I was very disappointed that the Demo did not work and that the documentation provided was inadequate for me to get it working.

So I will have to take much more time to get up to speed on CCS 7.0 and try to get the board working at a later date.

What I really wanted to do was examine the TI implementation of Time of Flight (TOF) using the Fast Fourier Transform (FFT) approach.  For those of you who are unfamiliar with an FFT, it takes the amplitude vs time data and extracts specific frequency artifacts hidden within the passive infrared (PIR) sensor response.

The TI researchers discovered that you can add another level of information extraction from the passive sensors by including information in the frequency domain.  IR emissive subjects, like humans and other mammals pass through the sensor field of view (FOV) at a relatively slow frequency at around 4-5 Hz.  A background of reflected sunlight would be at a lower frequency, while an active heat source like a fireplace generates a frequency signature above 50 Hz.  So the TOF sensor board samples at several hundred cycles per second to provide a good profile of likely heat sources within the sensor FOV.

The board itself provides two PIR sensors, an ambient light sensor and a humidity sensor.

The two PIR sensors provide a level of stereo viewing, which can be exploited to establish a level of three dimensional measurement.

The ambient light sensor provides data on the overall light contrast in the FOV.

The humidity sensor provides a reading on moisture content, which could affect the IR transmission from heat sources within the FOV.

The PIR sensors have a 65 degree FOV and are spaced ten centimeters apart, with a view intersection beginning about 15 cm from the board.

TI recommends that the board be placed so that the top PIR sensor is at about 1.8 meters above the floor, so that a human adult should trigger the top sensor first and smaller humans and mammals would trigger the lower sensor first.

Overall, the board provides the potential for a wealth of room/building surveillance opportunity for data mining and algorithm based data exploitation.

For my monitoring application, I have a fixed scene with a chair in which my subject (me) spends time on the laptop and viewing television.

So I can set up five monitoring states as follows:

     Waiting: The room is empty with the chair unoccupied.

     Transit: An IR source enters into the FOV of either PIR sensor.

     Good: An IR source sits in the chair and provides an equal intensity signal to both PIR sensors.

     Up:     A transition of IR detection from equal to predominately the upper PIR sensor, indicating that the subject has gotten out of the chair, probably to leave the room.

     Alert:     A transition of IR detection from equal to predominately the lower PIR sensor, indicating that the subject has fallen out of the chair and may need assistance.

Using the PIR intensity data, I can determine if the subject is approaching or moving away from the sensor board.

Over time, I can use the IR intensity data to do a rudimentary identification of the subject entering the FOV and sitting in the chair.

I can use the rate of change to identify if the subject is walking slowly or hurrying through the FOV.

After looking at these issues, it is clear that just this single board can be used for many purposes.

It could be used to monitor multiple subjects passing through a given area and build up unique identification profiles for each.  Sounds a little like the mine occupation/trespassing scenario Mehmet was looking at.

You could also use the sensor board to monitor livestock and even identify specific species within the building.  Could be useful for identifying predators and varmints.

From an Iot perspective, there are many different uses to which this board could be applied.

Now I have a question for TI:

Why did you include a 1.5v battery with a buck-boost converter to power the device when you will need a MCU to control and collect the data in close proximity?

Yes, you can, but what purpose does it serve.

Also, if you went to all of that trouble, then why did you not provide both a battery connection and BB output connection so that the MCU could monitor the battery and BB?

In remote locations, having a battery fuel gauge output would greatly assist support issues.

For monitoring humans, I would go with three PIR sensors so that you can get transit direction information in addition to high or low IR signals.

I would also sample the light sensor at the same frequency as the PIR sensors and build up a TOF profile on that data as well.  Having a frequency profile would provide even more useful exploitable information.

I am very happy with the board and I look forward to bringing it online and beginning to use it for my monitoring task.

I am also looking forward to learning how to exploit the wealth of information this one board can provide in a monitoring scenario.

Overall, I think TI did a great job, I just wish the demo had worked out of the box.

Thanks for reading,

DAB