search

cyberman54 / ESP32-Paxcounter

Wifi & BLE driven passenger flow metering with cheap ESP32 boards
https://cyberman54.github.io/ESP32-Paxcounter/
Other
1.69k stars 397 forks source link

BLE Scanner not working on Build 3.4.3 ? (TTGO LORA OLED BOARD 1.6.1) #921

Closed Qmick01 closed 1 year ago

Qmick01 commented 1 year ago

Hi, I am on build 3.4.3 and after flashing a few boards for BLE sensing only have found the BLE count is consistently 0.

I believe there may be an issue here and would appreciate any thoughts or a fix?

Paxcounter conf is below:

/ default settings for transmission of sensor data (first list = data on / second line = data off)

define PAYLOADMASK \

        ((GPS_DATA | MEMS_DATA | COUNT_DATA | SENSOR1_DATA | SENSOR2_DATA | SENSOR3_DATA | BATT_DATA) \
       & (~RESERVED_DATA))

// MAC sniffing parameters

define BLECOUNTER 1 // set to 0 if you do not want to install the BLE sniffer

define WIFICOUNTER 0 // set to 0 if you do not want to install the WIFI sniffer

define RSSILIMIT 0 // 0...-128, set to 0 if you do not want to filter signals

// BLE scan parameters

define BLESCANTIME 0 // [seconds] scan duration, 0 means infinite [default], see note below

define BLESCANWINDOW 80 // [milliseconds] scan window, see below, 3 .. 10240, default 80ms

define BLESCANINTERVAL 80 // [illiseconds] scan interval, see below, 3 .. 10240, default 80ms = 100% duty cycle

t-huyeng commented 1 year ago

Can you provide logs from the board?

FYI: you can format code blocks like this:

 code / conf etc. 
```
Qmick01 commented 1 year ago

Here are the logs for a few minutes of running. Is that helpful?

SERIAL PORT DATA

09:32:11.849 > 09:32:11.849 > rst:0x5 (DEEPSLEEP_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT) 09:32:11.853 > configsip: 188777542, SPIWP:0xee 09:32:11.853 > clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00 09:32:11.866 > mode:DIO, clock div:2 09:32:11.866 > load:0x3fff0030,len:1184 09:32:11.866 > load:0x40078000,len:13132 09:32:11.871 > load:0x40080400,len:3036 09:32:11.871 > entry 0x400805e4 09:32:11.930 > [ 20][I][configmanager.cpp:96] loadConfig(): [src/configmanager.cpp] Loading device configuration from NVRAM... 09:32:11.937 > [ 22][I][configmanager.cpp:122] loadConfig(): [src/configmanager.cpp] Runtime configuration v3.4.3 loaded 09:32:11.947 > [ 30][I][reset.cpp:72] do_after_reset(): [src/reset.cpp] Time spent in deep sleep: 600097 ms 09:32:11.958 > [ 38][I][main.cpp:115] setup(): [src/main.cpp] Starting paxcounter_7dd43646 v3.4.3 (runmode=2 / restarts=0) 09:32:11.965 > [ 47][I][main.cpp:116] setup(): [src/main.cpp] code build date: 1669277974 09:32:11.975 > [ 54][I][esp32-hal-i2c.c:75] i2cInit(): Initialising I2C Master: sda=21 scl=22 freq=100000 09:32:11.980 > [ 63][I][power.cpp:194] calibrate_voltage(): [src/power.cpp] ADC characterization based on reference voltage stored in eFuse 09:32:11.992 > [ 76][I][main.cpp:275] setup(): [src/main.cpp] Starting libpax... 09:32:11.997 > [ 79][I][main.cpp:285] setup(): [src/main.cpp] WIFISCAN: off 09:32:12.004 > [ 85][I][main.cpp:291] setup(): [src/main.cpp] BLESCAN: on 09:32:12.008 > [ 91][I][blescan.cpp:203] start_BLE_scan(): [bluetooth] Initializing bluetooth scanner ... 09:32:12.636 > [ 731][I][blescan.cpp:113] hci_cmd_send_ble_scan_start(): [bluetooth] BLE Scanning started 09:32:12.836 > [ 931][I][blescan.cpp:264] start_BLE_scan(): [bluetooth] Bluetooth scanner started 09:32:12.848 > [ 931][I][main.cpp:309] setup(): [src/main.cpp] Starting rcommand interpreter... 09:32:12.852 > [ 935][I][rcommand.cpp:515] rcmd_init(): [src/rcommand.cpp] Rcommand send queue created, size 55 Bytes 09:32:12.865 > [ 944][I][lorawan.cpp:248] lmic_init(): [src/lorawan.cpp] LORA send queue created, size 530 Bytes 09:32:12.874 > [ 969][I][lorawan.cpp:537] LoadLMICFromRTC(): [src/lorawan.cpp] LMIC state loaded 09:32:12.886 > [ 969][I][lorawan.cpp:293] lmic_init(): [src/lorawan.cpp] Already joined 09:32:12.891 > [ 972][I][lorawan.cpp:297] lmic_init(): [src/lorawan.cpp] Starting LMIC... 09:32:12.897 > [ 979][I][sdcard.cpp:95] sdcard_init(): [src/sdcard.cpp] looking for SD-card... 09:32:12.921 > E (1012) sdmmc_common: sdmmc_init_ocr: send_op_cond (1) returned 0x107 09:32:12.933 > E (1013) vfs_fat_sdmmc: sdmmc_card_init failed (0x107). 09:32:12.938 > [ 1017][E][sdcard.cpp:156] sdcard_init(): [src/sdcard.cpp] SD-card not found (263) 09:32:12.944 > E (1020) wifi_init: Failed to deinit Wi-Fi driver (0x3001) 09:32:12.950 > [ 1029][I][main.cpp:412] setup(): [src/main.cpp] Starting Interrupt Handler... 09:32:12.955 > [ 1036][I][main.cpp:440] setup(): [src/main.cpp] Starting Timers... 09:32:12.961 > [ 1042][I][timekeeper.cpp:391] time_init(): [src/timekeeper.cpp] Starting time pulse... 09:32:12.971 > [ 1051][I][timekeeper.cpp:194] timepulse_init(): [src/timekeeper.cpp] Timepulse: internal (ESP32 hardware timer) 09:32:12.978 > [ 1060][I][main.cpp:485] setup(): [src/main.cpp] Features: BATT OTA BLE LORA PACKED TIME 09:33:11.998 > [ 60096][I][reset.cpp:88] enter_deepsleep(): [src/reset.cpp] Preparing to sleep... 09:33:12.009 > [ 60096][I][blescan.cpp:277] stop_BLE_scan(): [bluetooth] Shutting down bluetooth scanner ... 09:33:12.015 > [ 60103][I][lorawan.cpp:219] lora_send(): [src/lorawan.cpp] 4 byte(s) sent to LORA 09:33:12.059 > [ 60157][I][blescan.cpp:286] stop_BLE_scan(): [bluetooth] Bluetooth scanner stopped 09:33:12.070 > [ 60157][I][reset.cpp:110] enter_deepsleep(): [src/reset.cpp] Waiting until send queues are empty... 09:33:18.200 > [ 66300][I][lorawan.cpp:219] lora_send(): [src/lorawan.cpp] 2 byte(s) sent to LORA 09:33:19.063 > [ 67162][I][reset.cpp:119] enter_deepsleep(): [src/reset.cpp] Waiting until LMIC is idle... 09:33:25.063 > [ 73162][I][lorawan.cpp:532] SaveLMICToRTC(): [src/lorawan.cpp] LMIC state saved 09:33:25.074 > [ 73163][I][reset.cpp:179] enter_deepsleep(): [src/reset.cpp] Going to sleep, good bye.

Qmick01 commented 1 year ago

I have just added to the thread.

On Thu, 24 Nov 2022, 17:39 t-huyeng, @.***> wrote:

Can you provide logs from the board?

— Reply to this email directly, view it on GitHub https://github.com/cyberman54/ESP32-Paxcounter/issues/921#issuecomment-1326727826, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOZZCO4VZRNEPTNQ5S6FLYTWJ6R5PANCNFSM6AAAAAASKF5VYM . You are receiving this because you authored the thread.Message ID: @.***>

cyberman54 commented 1 year ago

I can reproduce this. Will have to look deeper in this issue. Solution is probably to restructure code by removing compiler directives WIFICOUNTER and BLECOUNTER, and switch Wifi and BLE using runtime variables instead.

Qmick01 commented 1 year ago

Thank you. Would you know where this became an issue? I can report that version 3.3.2 does NOT have this bug and is working fine on sensors I have deployed.

cyberman54 commented 1 year ago

I compared 3.3.2 and 3.4.0, and did not find any relevant diffs in paxcounter code. But probably this could be causing this issue, is need to check further:

image

cyberman54 commented 1 year ago

This should be fixed now. Bumped new code as v3.4.4 /cc @Qmick01

Qmick01 commented 1 year ago

Hi,

I have been testing for 2 hours. I am now on 3.4.4 and have flashed two different versions of the same board, one a new type and one an older version I have stock of and used for many deployments. I am still unable to get a BLE number reported back. BLE 0 is all I get in several locations where other boards report BLE number. Also my tests revealed this build does not respond to an "80" on port 2 downlink.

cyberman54 commented 1 year ago

Then you probably use outdated code in your toolchain. Update Arduino Espressif32 framework and libraries of the project (i.e. libpax). Make sure that your settings in paxcounter.conf for WIFICOUNTER and BLECOUNTER are appropriate. Clear the NVRAM if board were flashed with other paxcounter build before.

Qmick01 commented 1 year ago

It's a brand new install of the code and the correct espressif framework came down.

I presume you used the same board and the new build worked?

On Sat, 26 Nov 2022, 21:25 Verkehrsrot, @.***> wrote:

Then you probably use outdated code in your toolchain. Update Arduino Espressif32 framework and libraries of the project (i.e. libpax). Make sure that your settings in paxcounter.conf for WIFICOUNTER and BLECOUNTER are appropriate. Clear the NVRAM if board were flashed with other paxcounter build before.

— Reply to this email directly, view it on GitHub https://github.com/cyberman54/ESP32-Paxcounter/issues/921#issuecomment-1328117043, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOZZCO3CE2CI7L7Q4DZM2J3WKJ555ANCNFSM6AAAAAASKF5VYM . You are receiving this because you were mentioned.Message ID: @.***>

Qmick01 commented 1 year ago

Here's the debug info:

But I keep getting 0 BLE.

23:37:48.814 > rst:0x5 (DEEPSLEEP_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT) 23:37:48.822 > configsip: 188777542, SPIWP:0xee 23:37:48.822 > clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00 23:37:48.828 > mode:DIO, clock div:2 23:37:48.828 > load:0x3fff0030,len:1184 23:37:48.828 > load:0x40078000,len:13132 23:37:48.833 > load:0x40080400,len:3036 23:37:48.833 > entry 0x400805e4 23:37:48.881 > [ 7][D][esp32-hal-cpu.c:244] setCpuFrequencyMhz(): PLL: 480 / 2 = 240 Mhz, APB: 80000000 Hz 23:37:48.889 > [ 21][I][configmanager.cpp:96] loadConfig(): [src/configmanager.cpp] Loading device configuration from NVRAM... 23:37:48.901 > [ 23][I][configmanager.cpp:122] loadConfig(): [src/configmanager.cpp] Runtime configuration v3.4.4 loaded 23:37:48.912 > [ 30][D][reset.cpp:51] do_after_reset(): [src/reset.cpp] Timezone set to CET-1CEST,M3.4.0/2,M10.4.0/3 23:37:48.923 > [ 39][I][reset.cpp:72] do_after_reset(): [src/reset.cpp] Time spent in deep sleep: 10105 ms 23:37:48.928 > [ 48][I][main.cpp:115] setup(): [src/main.cpp] Starting paxcounter_531f6c7e v3.4.4 (runmode=2 / restarts=0) 23:37:48.940 > [ 57][I][main.cpp:116] setup(): [src/main.cpp] code build date: 1669491922 23:37:48.945 > [ 64][I][esp32-hal-i2c.c:75] i2cInit(): Initialising I2C Master: sda=21 scl=22 freq=100000 23:37:48.956 > [ 73][I][power.cpp:194] calibrate_voltage(): [src/power.cpp] ADC characterization based on reference voltage stored in eFuse 23:37:48.967 > [ 87][I][main.cpp:274] setup(): [src/main.cpp] Starting libpax... 23:37:48.973 > [ 89][I][main.cpp:284] setup(): [src/main.cpp] WIFISCAN: off 23:37:48.978 > [ 95][I][main.cpp:290] setup(): [src/main.cpp] BLESCAN: on 23:37:48.984 > [ 101][I][blescan.cpp:203] start_BLE_scan(): [bluetooth] Initializing bluetooth scanner ... 23:37:49.620 > [ 751][I][blescan.cpp:113] hci_cmd_send_ble_scan_start(): [bluetooth] BLE Scanning started 23:37:49.826 > [ 952][I][blescan.cpp:264] start_BLE_scan(): [bluetooth] Bluetooth scanner started 23:37:49.866 > [ 952][I][main.cpp:300] setup(): [src/main.cpp] Starting rcommand interpreter... 23:37:49.866 > [ 956][I][rcommand.cpp:498] rcmd_init(): [src/rcommand.cpp] Rcommand send queue created, size 55 Bytes 23:37:49.866 > [ 965][I][lorawan.cpp:248] lmic_init(): [src/lorawan.cpp] LORA send queue created, size 530 Bytes 23:37:49.891 > [ 990][I][lorawan.cpp:537] LoadLMICFromRTC(): [src/lorawan.cpp] LMIC state loaded 23:37:49.891 > [ 990][I][lorawan.cpp:293] lmic_init(): [src/lorawan.cpp] Already joined 23:37:49.891 > [ 993][I][lorawan.cpp:297] lmic_init(): [src/lorawan.cpp] Starting LMIC... 23:37:49.891 > [ 1000][I][sdcard.cpp:95] sdcard_init(): [src/sdcard.cpp] looking for SD-card... 23:37:49.906 > E (1033) sdmmc_common: sdmmc_init_ocr: send_op_cond (1) returned 0x107 23:37:49.917 > E (1034) vfs_fat_sdmmc: sdmmc_card_init failed (0x107). 23:37:49.923 > [ 1038][E][sdcard.cpp:156] sdcard_init(): [src/sdcard.cpp] SD-card not found (263) 23:37:49.928 > [ 1045][I][main.cpp:396] setup(): [src/main.cpp] Starting Interrupt Handler... 23:37:49.934 > [ 1052][I][main.cpp:424] setup(): [src/main.cpp] Starting Timers... 23:37:49.939 > [ 1058][I][timekeeper.cpp:391] time_init(): [src/timekeeper.cpp] Starting time pulse... 23:37:49.950 > [ 1067][I][timekeeper.cpp:194] timepulse_init(): [src/timekeeper.cpp] Timepulse: internal (ESP32 hardware timer) 23:37:49.956 > [ 1076][I][main.cpp:469] setup(): [src/main.cpp] Features: BATT OTA LORA PACKED TIME 23:38:19.950 > [ 31076][D][cyclic.cpp:23] doHousekeeping(): [src/cyclic.cpp] Heap: Free:159204, Min:150512, Size:224708, Alloc:110580, StackHWM:3536 23:38:19.961 > [ 31077][D][cyclic.cpp:28] doHousekeeping(): [src/cyclic.cpp] IRQhandler 2708 bytes left | Taskstate = 0 23:38:19.968 > [ 31086][D][cyclic.cpp:31] doHousekeeping(): [src/cyclic.cpp] Rcommand interpreter 2524 bytes left | Taskstate = 2 23:38:19.978 > [ 31096][D][cyclic.cpp:36] doHousekeeping(): [src/cyclic.cpp] LMiCtask 3540 bytes left | Taskstate = 1 23:38:19.989 > [ 31105][D][cyclic.cpp:40] doHousekeeping(): [src/cyclic.cpp] Lorasendtask 2460 bytes left | Taskstate = 2 23:38:48.968 > [ 60100][D][senddata.cpp:74] sendData(): [src/senddata.cpp] Sending count results: pax=0 / wifi=0 / ble=0 23:38:48.983 > [ 60104][D][senddata.cpp:15] SendPayload(): [src/senddata.cpp] sending Payload for Port 1 23:38:48.990 > [ 60109][D][senddata.cpp:15] SendPayload(): [src/senddata.cpp] sending Payload for Port 8 23:38:49.004 > [ 60115][I][reset.cpp:88] enter_deepsleep(): [src/reset.cpp] Preparing to sleep... 23:38:49.004 > [ 60122][I][blescan.cpp:277] stop_BLE_scan(): [bluetooth] Shutting down bluetooth scanner ... 23:38:49.014 > [ 60132][D][lorawan.cpp:400] myEventCallback(): [src/lorawan.cpp] TXSTART 2 23:38:49.018 > [ 60138][I][lorawan.cpp:219] lora_send(): [src/lorawan.cpp] 4 byte(s) sent to LORA 23:38:49.024 > [ 60147][I][blescan.cpp:286] stop_BLE_scan(): [bluetooth] Bluetooth scanner stopped 23:38:49.035 > [ 60152][I][reset.cpp:110] enter_deepsleep(): [src/reset.cpp] Waiting until send queues are empty... 23:38:54.661 > [ 65788][D][lorawan.cpp:400] myEventCallback(): [src/lorawan.cpp] RXSTART 1 23:38:55.661 > [ 66787][D][lorawan.cpp:400] myEventCallback(): [src/lorawan.cpp] RXSTART 1 23:38:55.698 > [ 66829][D][lorawan.cpp:400] myEventCallback(): [src/lorawan.cpp] TXCOMPLETE 1 23:38:56.409 > [ 67541][I][lorawan.cpp:219] lora_send(): [src/lorawan.cpp] 2 byte(s) sent to LORA 23:38:57.034 > [ 68161][I][reset.cpp:119] enter_deepsleep(): [src/reset.cpp] Waiting until LMIC is idle... 23:39:54.946 > [126073][D][lorawan.cpp:400] myEventCallback(): [src/lorawan.cpp] TXSTART 23:40:00.590 > [131722][D][lorawan.cpp:400] myEventCallback(): [src/lorawan.cpp] RXSTART 23:40:01.595 > [132722][D][lorawan.cpp:400] myEventCallback(): [src/lorawan.cpp] RXSTART 23:40:01.631 > [132764][D][lorawan.cpp:400] myEventCallback(): [src/lorawan.cpp] TXCOMPLETE 23:40:02.029 > [133161][I][lorawan.cpp:532] SaveLMICToRTC(): [src/lorawan.cpp] LMIC state saved 23:40:02.041 > [133162][I][reset.cpp:179] enter_deepsleep(): [src/reset.cpp] Going to sleep, good bye. 23:40:11.983 > ets Jun 8 2016 00:22:57 23:40:11.988 > Michael Joseph +44 (0) 7790928628 "In Velo, Veritas"

On Sat, Nov 26, 2022 at 9:25 PM Verkehrsrot @.***> wrote:

Then you probably use outdated code in your toolchain. Update Arduino Espressif32 framework and libraries of the project (i.e. libpax). Make sure that your settings in paxcounter.conf for WIFICOUNTER and BLECOUNTER are appropriate. Clear the NVRAM if board were flashed with other paxcounter build before.

— Reply to this email directly, view it on GitHub https://github.com/cyberman54/ESP32-Paxcounter/issues/921#issuecomment-1328117043, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOZZCO3CE2CI7L7Q4DZM2J3WKJ555ANCNFSM6AAAAAASKF5VYM . You are receiving this because you were mentioned.Message ID: @.***>

cyberman54 commented 1 year ago

please post platformio.ini and paxconter.conf of your build.

cyberman54 commented 1 year ago

Make sure you're using current libpax lib, look in the build processing notes:

Scanning dependencies... Dependency Graph |-- libpax @ 1.0.2+sha.2f774b5

Qmick01 commented 1 year ago

Thanks.

I have assumed a new install will automatically pull the right versions of everything down and install but perhaps not.

Does the downlink work for you on version 3.4.4?

I couldn't get a response.

I will have a look at the version log later.

On Sun, 27 Nov 2022, 11:09 Verkehrsrot, @.***> wrote:

Make sure you're using current libpax lib, look in the build processing notes:

Scanning dependencies... Dependency Graph |-- libpax @ 1.0.2+sha.2f774b5

— Reply to this email directly, view it on GitHub https://github.com/cyberman54/ESP32-Paxcounter/issues/921#issuecomment-1328223006, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOZZCO4QRGRIAJLBNWH7ZN3WKM6O7ANCNFSM6AAAAAASKF5VYM . You are receiving this because you were mentioned.Message ID: @.***>

cyberman54 commented 1 year ago

I don't see any problems with downlinks. Everything is working, as it did before.

Did you copy platformio_orig.ini to your platformio.ini when upgrading your build installation? If not, you're probably using outdated / incompatible libs.

Qmick01 commented 1 year ago

Ok. Thanks for checking. I will do some more testing.

On Sun, 27 Nov 2022, 12:00 Verkehrsrot, @.***> wrote:

I don't see any problems with downlinks. Everything is working, as it did before.

— Reply to this email directly, view it on GitHub https://github.com/cyberman54/ESP32-Paxcounter/issues/921#issuecomment-1328231013, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOZZCOYLUPQDSQTSHJBZBHTWKNEMNANCNFSM6AAAAAASKF5VYM . You are receiving this because you were mentioned.Message ID: @.***>

cyberman54 commented 1 year ago

@Qmick01 for debugging, you can use new feature logging system output to SD card

cyberman54 commented 1 year ago

Received by mail from @Qmick01:

I have just re-cloned paxcounter on a different machine. It does compile and everything looks great - but still no BLE count.

I need to flash some more boards this week and have been using a version from February as this works reliably for most things.

I attach the compile output - perhaps there is something in there I am not seeing?

Can not remove temporary directory C:\Users\Mike\Documents\PlatformIO\Projects\3.4.4 new\ESP32-Paxcounter\.pio\build. Please remove it manually to avoid build issues Processing usb (framework: arduino; board: esp32dev; platform: espressif32@5.2.0) ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Verbose mode can be enabled via -v, --verbose option Parsing hardware configuration from C:\Users\Mike\Documents\PlatformIO\Projects\3.4.4 new\ESP32-Paxcounter\src\hal\ttgov21new.h Parsing LMIC configuration from C:\Users\Mike\Documents\PlatformIO\Projects\3.4.4 new\ESP32-Paxcounter\src\lmic_config.h Parsing LORAWAN keys from C:\Users\Mike\Documents\PlatformIO\Projects\3.4.4 new\ESP32-Paxcounter\src\loraconf.h Parsing OTA keys from C:\Users\Mike\Documents\PlatformIO\Projects\3.4.4 new\ESP32-Paxcounter\src\ota.conf Target board: esp32dev @ 921600bps Partition table: min_spiffs.csv CONFIGURATION: https://docs.platformio.org/page/boards/espressif32/esp32dev.html PLATFORM: Espressif 32 (5.2.0) > Espressif ESP32 Dev Module HARDWARE: ESP32 240MHz, 320KB RAM, 4MB Flash DEBUG: Current (cmsis-dap) External (cmsis-dap, esp-bridge, esp-prog, iot-bus-jtag, jlink, minimodule, olimex-arm-usb-ocd, olimex-arm-usb-ocd-h, olimex-arm-usb-tiny-h, olimex-jtag-tiny, tumpa) PACKAGES:

Leaving... Hard resetting via RTS pin... ===================================================================================== [SUCCESS] Took 73.41 seconds =====================================================================================

Environment Status Duration

usb SUCCESS 00:01:13.414 ====================================================================================== 1 succeeded in 00:01:13.414 ====================================================================================== * Terminal will be reused by tasks, press any key to close it.

cyberman54 commented 1 year ago

The toolchain seems current, i checked the compiler log.

I cannot reproduce the issue. @t-huyeng @spmrider could you, please, test the current code and look if you can reproduce this issue? Thanks.

cyberman54 commented 1 year ago

@Qmick01 please look on the display, while running v3.4.4 - what is shown in line with "RLIM:" ?

t-huyeng commented 1 year ago

I can not reproduce this issue. Looks like everything is working fine.

I also cloned the repository completed new and used a new board (Heltec). See my configs here.

My steps as descript in the documentations after cloning:


➜  ESP32-Paxcounter  (master) ✔ cp platformio_orig.ini platformio.ini

➜  ESP32-Paxcounter  (master) ✔ cp src/paxcounter_orig.conf src/paxcounter.conf

➜  ESP32-Paxcounter  (master) ✔ cp src/loraconf_sample.h src/loraconf.h

➜  ESP32-Paxcounter  (master) ✔ cp src/ota_sample.conf src/ota.conf

I only updated paxcounter.conf to enable BLECOUNTER see here and set my board to heltecv21.h in the platformio.ini see here .

I also updated the lorawan keys loraconf.h.

Running Heltec board: image

Grafana Dashboard: image

Qmick01 commented 1 year ago

So this morning the 3.4.4 board is showing a ble count of 1 but two units on an older build are running next to it and they both show a ble count of 5.

The 3.4.4 screen shots are below. It shows RSSI 37 and RLIM off.

I don't set the RSSI in my config file so that is whatever it is defaulted to. See image of the conf file.

If that's the actual RSSI then it could be a problem. What should I set it to in the conf file so it defaults to no limit?

Perhaps this is the issue.

20221202_093051 Screenshot_20221202_092724_Gallery Screenshot_20221202_093251_Gallery

t-huyeng commented 1 year ago

@Qmick01 I think your config file picture is missing. You are also sharing some private pictures with your pictures currently.

It is also better to share your config file as formatted code as I described https://github.com/cyberman54/ESP32-Paxcounter/issues/921#issuecomment-1326727826 here or as a gist.

Qmick01 commented 1 year ago

I have added the photo of the Paxcounter.conf section showing the RSSILIMIT value at zero.

Any thoughts?

t-huyeng commented 1 year ago

Have you read the comment behind #define RSSILIMIT 0 ?

It states this: // 0...-128, set to 0 if you do not want to filter signals So if you have it set to 0 (default) no signal filtering will happen.

It would also be much easier to help you if you could share config files as formatted code as I described https://github.com/cyberman54/ESP32-Paxcounter/issues/921#issuecomment-1326727826 here or as a gist.

I also tried my board with #define WIFICOUNTER 0 and it works fine. I would suggest you check again if your files are like this https://gist.github.com/t-huyeng/42e08a22ca2a7f71ea00f681dded2d9f

How old is the version you are running on board which are showing more BLE-Counts? Which version of libpax is running there?

As far as I know there has been changes to the libpax for better filtering (maybe this here https://github.com/dbSuS/libpax/commit/277a5d2959512f33c6aac74b9eed0f002673f274).

cyberman54 commented 1 year ago

As far as I know there has been changes to the libpax for better filtering (maybe this here >https://github.com/dbSuS/libpax/commit/277a5d2959512f33c6aac74b9eed0f002673f274).

Exactly. Thus, count values generated with older versions are not comparable with newer ones.

Qmick01 commented 1 year ago

Yes. No RSSILIMIT value means to filtering on RSSI. That's what I would expect.

The other boards are running 3.1.2 version of the code. Is the new LIBPAX subsequent to that version?

Qmick01 commented 1 year ago

Thanks for this.

That's a very significant change as it means historical data no longer lines up with current data so making a timeline of activity over time becomes quite challenging as the meaning of changing pax values could be mixed up between the software changes and the actual pax values.

Is there a clear sense of what the change is based on?

For example, was the change made to eliminate a specific issue or count problem?

cyberman54 commented 1 year ago

This change is now already ~1 year in place. Reasons were 1) improving privacy 2) studies showed that using randomized MACs gives better results on number of pax, because headless devices like cars, scooters etc. are better filtered this way

Qmick01 commented 1 year ago

Thanks for help figuring this out.

I have had cases where this build sees no BLE even when I am in the room with my phone BLE in the same room. This morning it saw one BLE when the others saw 5 and no doubt there is evidence as to why either number is correct.

I will do some more testing.

cyberman54 commented 1 year ago

Remember that libpax lib is an approach to estimate crowds and/or relative changes over time. Not single persons, not absolute values.

Qmick01 commented 1 year ago

Yes, that's how I have been using it, and often having some challenges getting numbers that map to a rough headcount so if a change in the algorithm makes this better then that's great news.

Do you have any insights as to how well the new count numbers equate to actual headcount?