Grove AI not appearing as a drive on MacOS

My new Grove Vision AI sensor won’t appear as a drive in Finder on my Mac. Double-clicking the BOOT button doesn’t resolve the issue and neither does switching USB-C cables.

Chrome gives me a “Grove AI detected” pop-up, however, so it is able to detect the sensor.

I’m unable to follow the Bouffalo Lab DevCube steps, because they don’t seem to offer a Mac version. (And seems like that shouldn’t be necessary since the firmware should already be up to date on a new sensor.)

Would greatly appreciate any help. Thank you!!

Hi,
image
I’m sorry that seeedstudio only provides bootloader firmware for windows, and I’m sorry that it can’t be recognized on MacOS, since everyone in Seeedstudio uses windows.

Ah, gotcha. I’ll have to see if I can use a friend’s PC for this, then. Thank you for the quick response!

Hi again- Hoping you can help me further troubleshoot my Grove Vision AI sensor. (And hoping I didn’t somehow brick it!!!)

  • I’m using a Windows 11 PC now.
  • The GROVEAI appears as a drive but I get an error (with no details) when I try to copy over the .uf2 file with my custom model. (It’s about 22MB.)
  • The serial monitor fails to display the message shown in the tutorial.
  • I installed Bouffalo Lab Dev Cube 1.8.5 but am unable to update the bootloader. I get this error when trying to Create & Download the tinyuf2-grove_visiono_ai_v2.0.1.bin file:

[17:16:01.234] - Close COM3 Success

[17:16:01.255] - EFUSE_CFG

[17:16:01.256] - BOOTHEADER_CFG

[17:16:01.271] - Create bootheader using C:\Users\olive\Downloads\BouffaloLabDevCube-v1.8.5\chips\bl702\img_create_mcu/efuse_bootheader_cfg.ini

[17:16:01.272] - Updating data according to <C:\Users\olive\Downloads\BouffaloLabDevCube-v1.8.5\chips\bl702\img_create_mcu/efuse_bootheader_cfg.ini[BOOTHEADER_CFG]>

[17:16:01.275] - Created file len:176

[17:16:01.279] - Create efuse using C:\Users\olive\Downloads\BouffaloLabDevCube-v1.8.5\chips\bl702\img_create_mcu/efuse_bootheader_cfg.ini

[17:16:01.282] - Updating data according to <C:\Users\olive\Downloads\BouffaloLabDevCube-v1.8.5\chips\bl702\img_create_mcu/efuse_bootheader_cfg.ini[EFUSE_CFG]>

[17:16:01.286] - Created file len:128

[17:16:01.292] - Image create path: C:\Users\olive\Downloads\BouffaloLabDevCube-v1.8.5\chips\bl702\img_create_mcu

[17:16:01.293] - Config file: C:\Users\olive\Downloads\BouffaloLabDevCube-v1.8.5\chips\bl702\img_create_mcu/img_create_cfg.ini

[17:16:01.297] - Image hash is b’478789c81acd02b2d0ff85bbe4c720c0a49b3dbfbe04f77e3e984376384bcfaa’

[17:16:01.299] - Header crc: b’75e9f079’

[17:16:01.299] - Write flash img

[17:16:01.301] - Encrypt efuse data

[17:16:01.304] - ========= eflash loader config =========

[17:16:01.378] - Version: eflash_loader_v2.4.6

[17:16:01.379] - Program Start

[17:16:01.380] - ========= eflash loader cmd arguments =========

[17:16:01.381] - None

[17:16:01.381] - Config file: C:\Users\olive\Downloads\BouffaloLabDevCube-v1.8.5\chips\bl702\eflash_loader\eflash_loader_cfg.ini

[17:16:01.383] - serial port is COM3

[17:16:01.383] - cpu_reset=False

[17:16:01.385] - chiptype: bl702

[17:16:01.386] - ========= Interface is uart =========

[17:16:01.388] - com speed: 2000000

[17:16:01.389] - Eflash load helper file: C:\Users\olive\Downloads\BouffaloLabDevCube-v1.8.5\chips\bl702\eflash_loader/eflash_loader_32m.bin

[17:16:01.390] - ========= load eflash_loader.bin =========

[17:16:01.390] - Load eflash_loader.bin via uart

[17:16:01.391] - ========= image load =========

[17:16:01.500] - ack is

[17:16:01.501] - Not ack OK

[17:16:01.501] - result:

[17:16:01.782] - tx rx and power off, press the machine!

[17:16:01.785] - cutoff time is 0.05

[17:16:01.844] - power on tx and rx

[17:16:02.847] - reset cnt: 0, reset hold: 0.05, shake hand delay: 0.1

[17:16:02.848] - clean buf

[17:16:02.858] - send sync

[17:20:22.311] - Error: Write timeout

[17:20:22.313] - shake hand fail

[17:20:22.314] - ########################################################################

[17:20:22.315] - 请按照以下描述排查问题:

[17:20:22.316] - GPIO28是否上拉到板子自身的3.3V,而不是外部的3.3V

[17:20:22.316] - GPIO15(RX)是否连接到USB转串口的TX引脚

[17:20:22.317] - GPIO14(TX)是否连接到USB转串口的RX引脚

[17:20:22.317] - 在使用烧录软件进行烧录前,是否在GPIO28拉高的情况下,使用Reset/Chip_En复位了芯片

[17:20:22.318] - 烧录软件所选择的COM口,是否是连接芯片的串口

[17:20:22.318] - 烧录软件上选择的波特率是否是USB转串口支持的波特率

[17:20:22.319] - 3.3V供电是否正常

[17:20:22.320] - 板子供电电流是否正常(烧录模式下,芯片耗电电流5-7mA)

[17:20:22.322] - ########################################################################

[17:20:22.323] - shake hand fail

[17:20:22.346] - Load helper bin time cost(ms): 260955.73120117188

[17:20:22.347] - ErrorCode: 0050, ErrorMsg: BFLB IMG LOAD SHAKEHAND FAIL

[17:20:22.347] - ErrorCode: 0003, ErrorMsg: BFLB LOAD HELP BIN FAIL

[17:20:22.348] - Burn Retry

[17:20:22.348] - 0

[17:20:22.348] - Burn return with retry fail

Hi, do you follow the content about custom training model in the wiki: Train and Deploy Your Own AI Model with Roboflow, YOLOv5, TensorFlow Lite | Seeed Studio Wiki

In addition, I saw a lot of error prompts in the error message, and you should check it. Is there any problem.Another question is whether you refresh firmware?
image

I seem to be unable to refresh the firmware. When I run the sketch below, the serial port gives me an “Algo begin failed.” message and it doesn’t erase the firmware.

I’m using an Uno with a Grove shield. I’ve tried resetting/using different cables, etc.

Please advise if you can. Thanks!

#include “Seeed_Arduino_GroveAI.h”
#include <Wire.h>
WEI wei(Wire);

void setup()
{
Wire.begin();
Serial.begin(115200);
while (!Serial)
{
/* code */
}
delay(1000);
Serial.println(“Press any key to start erase”);
}

void loop()
{
if (Serial.available())
{
while (Serial.available())
{
Serial.read();
}
uint32_t id = wei.ID();
Serial.print(“ID: 0x”);
Serial.println(id, HEX);
wei.enter_bootloader();
Serial.println(“Erase”);
wei.erase();
Serial.println(“Erase done”);
}
}

Hello, have you remove the external device when you restore the ai board firmware? “Algo begin failed” is the output from our example, could you click the reset button first and follow the tutorial here to try again? Grove - Vision AI Module | Seeed Studio Wiki

Interestingly, Chrome does give me a “Grove AI detected” pop-up, indicating that the sensor is being detected by the browser. I’m unable to follow the Bouffalo Lab DevCube steps, as they don’t seem to offer a Mac version. Given that this is a new sensor, I assume the firmware should already be up to date.