AWR1642BOOST: Extending Range of Vehicle Occupancy Demo

You can use the sensing estimator tool to find out profile configuration which you need to feed into \chirp_configs\od_demo_car.cfg file at profileCfg command. Refer mmWave SDK User guide to find out each parameter definition of profileCfg CLI command to convert sensing-estimator-tool info to CLI based command.

Regards,

Jitendra

Hi Andrew,

What is the range you are interested in?  As Jiten mentioned, you will need to modify the profileCfg parameters.  Specifically, the number of samples and sampling time will need to be increased while maintaining the same overall chirp time. Which means a shorter inter-chirp period for the 1D FFT to run.  The current demo and GUI code are hard coded to 64 range rows. This will also need to be increased. The problem here is the amount of data to be sent over UART will grow.  One solution to this is to convert the heatmap to 16 or 8-bit prior to transmission.

  -dave

I would like to extend the range out to 30-50 meters. The current frame rate appears to be around 6 fps. I am fine with slowing this down to 1fps or less if necessary to accommodate data transfer over UART.

I am concerned about running out of memory for storing ADC samples and the final radar data cube. The AWR1642 has 1.5 Mb of memory. Are all the ADC samples saved until the end of the frame, or are they overwritten after the 1D FFT is complete?

I will look into your suggested changes to the chirp profile.

Thanks,

Andrew

I tried saving a chirp configuration using the mmWave Demo Visualizer and by inputting the values from the Sensing estimator into the profileCfg command. I have been playing around with the parameters because I think the sensor is running out of memeory.

The VOD demo collects more data than a standard mmWave chirp configuration. I have included one of the default chirp configurations below. (od_demo_car_0317_1p0.cfg from the Automotive Toolbox version 2.4.3)

Both transmitters are active twice per loop rather than only once. How does this affect the data processing? Are there other considerations that must be made in order to have a valid chirp profile?

sensorStop

flushCfg
dfeDataOutputMode 1
channelCfg 15 3 0
adcCfg 2 1
adcbufCfg -1 0 0 1 1
profileCfg 0 77 250 10 40 0 0 98 1 64 2200 0 0 40
chirpCfg 0 0 0 0 0 0 0 1
chirpCfg 1 1 0 0 0 0 0 2
chirpCfg 2 2 0 0 0 0 0 1
chirpCfg 3 3 0 0 0 0 0 2
frameCfg 0 3 128 0 160 1 0
lowPower 0 1
guiMonitor -1 0 1 1
calibDcRangeSig -1 0 -5 8 256

%OD Demo commands:
zoneDef 2 13 16 9 15 13 16 27 15
coeffMatrixRow 0 -14.409613 -8.187467 -8.019457 3.833826 4.045485 -0.539210
coeffMatrixRow 1 -12.465002 -2.228748 -5.612140 4.324446 -7.912656 0.528747
coeffMatrixRow 2 -9.210626 -3.004206 -0.082298 -2.949700 7.538751 0.246658
coeffMatrixRow 3 -1.917373 2.299849 2.389683 7.035545 6.675194 -0.099186
meanVector 23.992751 23.689970 -3.388539 -3.691319 0.816150
stdVector 8.495360 8.497394 2.198640 2.432388 0.196337
oddemoParms 12 0.001
sensorStart

The VOD range resolution is 4.69cm, which gives a max range of 3m with 64 range rows. 30m would require 640 range rows, which will simply not fit into memory. All 1D FFT outputs are stored in the radar cube for frame level processing, and the current cube size is 512KB.

The goal of the VOD demo is to detect very fine movements, primarily breathing, which is probably more appropriate for smaller ranges. I suspect that for such a large range, your goal is simply to detect objects/people moving into and out of monitored zones. We have another demo for this in the TIRex Industrial Toolbox. In TIRex, navigate to Software -> mmWave Sensors -> Industrial Toolbox -> Labs -> Zone Occupancy Detection. It does support AWR1642, but there's a bug in the currently released version (that will be updated in early Dec). LMK if you'd like to try it, I can put a version out on Box for you.

--dave

I would like to get post-processed data out of the sensor. It doesn't need to be the full radar cube, but the Range-Azimuth Heatmap shown in the mmWave Demo Visualizer would be ideal if it wasn't only for the 0-doppler bin.

A range of 30-50m and a range resolution of 50cm. would be acceptable This looks like a valid chirp configuration in the Sensing Estimator. I haven't had any success by changing the chirp profile in the VOD demo. Should I keep trying to get this working, or is there a better demo to get a Heatmap.

The two options I considered were the Range-Azimuth Heatmap published by the mmWave Demo Visualizer and the Vehicle Occupancy Demo Heatmap. I chose the VOD demo because the Heatmap shows data from all velocity bins instead of only the 0-doppler bin.

I believe the Zone Occupancy Demo only shows point cloud data.

I checked with the engineers that designed this chirp configuration and to go out to 30m isn't feasible for memory and timing reasons. They recommended other demos such as People Counting, or Zone Occupancy. There is an Obstacle Detection (ODS) demo that can generate a heatmap out to 20m, but I think it is zero doppler only.

I have tried the Obstacle Detection demo from (lab0004 from Automotive Toolbox version 2.4.3), and it doesn't seem to have a heatmap. Is there a setting to enable it?

Thank you,

Check the guiMonitor command described in the user guide. There are two that can be enabled, one at a time. And you will need to lower the frame rate to 4fps, lower if you increase the range. The range is increased via the last parameter of the cfarCfg command. Note, that this demo is designed for an EVM with an antenna pattern that creates a different virtual antenna, so objects in the heatmaps may not track as you expect.