Release type: Feature Release
Date: July 22, 2024
- AAGBT-170: Added additional AT commands when setting up MQTT over TLS to allow the device to pass the AWS IoT Core certification tests.
- MQTT3.1.1
- Set SNI configuration
- TLS authorization to required
None
- AAGBT-165: When capturing IMU data converted from ADC values to accelerometer and gyro units to debug sometimes the application hangs. This appears to be a floating point printf() issue and is currently masked by reducing the floating point precision to 3 places. For full precision data, please capture converted IMU data to the microSD card file system.
- AAGBT-159: The EVK does not power up with powered by just a battery. See troubleshooting guide for a work around
- AAGBT-159: When powering EVK from a battery the RGB LEDs are not powered. See troubleshooting guide for a hardware work around
- AAGBT-105: When using a custom BLE name, the OOB application is not able to connect to the Python GUI demo application
- AAGBT-123: When capturing audio, or IMU data if there is an issue receiving data from the NDP120 the application hangs
Release type: Feature Release
Date: June 25, 2024
- AAGBT-166: Add config.ini support to identify how the NDP120 pulls in data for each model (audio, IMU, or audio & IMU)
- AAGBT-168: Add config.ini support to identify which LED color lights by neural network number and inference index
None
- AAGBT-165: When capturing IMU data converted from ADC values to accelerometer and gyro units to debug sometimes the application hangs. This appears to be a floating point printf() issue and is currently masked by reducing the floating point precision to 3 places. For full precision data, please capture converted IMU data to the microSD card file system.
- AAGBT-159: The EVK does not power up with powered by just a battery. See troubleshooting guide for a work around
- AAGBT-159: When powering EVK from a battery the RGB LEDs are not powered. See troubleshooting guide for a hardware work around
- AAGBT-105: When using a custom BLE name, the OOB application is not able to connect to the Python GUI demo application
- AAGBT-123: When capturing audio, or IMU data if there is an issue receiving data from the NDP120 the application hangs
Release type: Feature Release
Date: June 20, 2024
- AAGBT-161: Update the OOB application to use the Syntiant SDK V110. This update was required to support running multiple ML models on the NDP120.
- AAGBT-161: Added new NDP120 configuration "Mode=8" that runs the circular motion model and the 5-keyword model at the same time
- AAGBT-161: Updated the AWS and IoTConnect AWS configurations to capture and send telemetry data for multiple models running on the NDP120 at the same time
- AAGBT-160: Updated IMU *.csv header labels/headers to match the strings used in Edge Impulse generated data captures
- AAGBT-162: Added the ability to convert IMU ADC values to accelerometer/gyro floating point data
- AAGBT-164: Added a configuration item to disable the "down" "down" sequence to enter low-power mode. The default behavior is unchanged
None
- AAGBT-165: When capturing IMU data converted from ADC values to accelerometer and gyro units to debug sometimes the application hangs. This appears to be a floating point printf() issue and is currently masked by reducing the floating point precision to 3 places. For full precision data, please capture converted IMU data to the microSD card file system.
- AAGBT-159: The EVK does not power up with powered by just a battery. See troubleshooting guide for a work around
- AAGBT-159: When powering EVK from a battery the RGB LEDs are not powered. See troubleshooting guide for a hardware work around
- AAGBT-105: When using a custom BLE name, the OOB application is not able to connect to the Python GUI demo application
- AAGBT-123: When capturing audio, or IMU data if there is an issue receiving data from the NDP120 the application hangs
Release type: Documentation Release
Date: Feb 28, 2024
Note: The V1.5.1 release is compatible with Edge Impulse Models generated between 2/7/2024 and June 9th 2024.
- AAGBT-143: Update IoTConnect documentation. Add new device and dashboard templates.
- AAGBT-145: Add new document RASynEdgeImpulseDataIngestion.md detailing how to use the Edge Impulse Daemon to capture training data from the RASynBoard directly into Edge Impulse
- AAGBT-146: Add a video link showcasing capturing training data using the Edge Impulse Daemon
- AAGBT-147: Update the troubleshooting guide to include common Renesas Flash Programmer errors
None
- AABGT-105: When using a custom BLE name, the OOB application is not able to connect to the Python GUI demo application
- AABGT-123: When capturing audio, or IMU data if there is an issue receiving data from the NDP120 the application hangs
Release type: Feature Release
Date: Jan 29, 2024
- AAGBT-42: Add AWS IoT Core support
- AAGBT-76: Add a developers guide to help developers leverage the OOB application for custom designs
- AABGT-114: Add support to modify the decimation_inshift model values from the config.ini file
- AABGT-117: Add Syntiant SDK v105 support
- AABGT-126: Create a video series to help developers use the OOB application
- AABGT-129: Create a troubleshooting guide to help developers overcome common issues
- AABGT-130: Update the application to capture and report inference count by inference ID. Add this data to any cloud connected telemetry messages.
- AABGT-142: Add new
[Function_x]blocks targeting the Edge Impulse motion and audio models to allow users to drop in models generated on Edge Impulse. - AABGT-112: Add a time server configuration item to the config.ini file and underlying code to support using a custom time server
- AAGBT-82: Corrected an issue where the circular IMU model could not capture IMU data for training
- AABGT-120: Corrected an issue where the device would not get an IP address when using the Renesas WiFi Configuration mobile device application
- AABGT-113: Improved debug when the application does not get a valid IP address or time from the time server
- AABGT-104: Corrected an issue where the dual microphone models did not work
- AABGT-105: When using a custom BLE name, the OOB application is not able to connect to the Python GUI demo application
- AABGT-123: When capturing audio, or IMU data if there is an issue receiving data from the NDP120 the application hangs
Release type: Feature Release
Date: Oct 26, 2023
- AAGBT-98: Updated the Syntiant SDK to v103 for Edge Impulse compatibility; this change includes updated *.synpkg files.
- AAGBT-102: Updated IoTConnect documentation to include links to the quick start guide
- AAGBT-96: Only pull the DA16600 out of reset if it's needed for BLE or Cloud connectivity features
- AAGBT-93: Add configuration item
[WIFI]--> Use_Config_AP_Detailsto allow use of the Renesas iOS or Android WiFi Configuration application - AAGBT-110: Add serial command
logto dump early Syntiant debug when using the com port enumerated on the core board UsB-C port - AAGBT-111: Add new config.ini function block
[Function_4]to support ML models generated by Edge Impulse
- AAGBT-79: Modified the IoTConnect logic after the device enters/exits low power mode. The application now allows the DA16600 MQTT logic to re-establish the MQTT connection.
- AAGBT-92: Added logic to save all configuration items from the config.ini file into SPI Flash when flashing models using the user button press > 3 seconds.
- AABGT-105: When using a custom BLE name, the OOB application is not able to connect to the Python GUI demo application
Release type: Feature Release f Date: Sept 1, 2023
- AABGT-43: Added logic to connect to Avnet's IoTConnect and send inference event details as telemetry
- AABGT-44: Added functionality to store Wi-Wi SSID and password in config.ini file
- AAGBT-75: Added documentation on all the OOB application configuration items and features
- AAGBT-80: Added IoTConnect documentation
- AAGBT-85: Added functionality to store the contents of the config.ini file to SPI flash when the user presses the user button for > 3 seconds. Additionally when the device boots without the microSD card present, the configuration is read from SPI Flash.
- AABGT-86: Added functionality to read AWS certificates from the microSD card
- AAGBT-90: Added documentation on how to load one of the OOB releases without the Renesas e^2 studio
- AABGT-84: Corrected an issue where BLE inference data was not sent when running the core board in stand alone mode. Corrected by AABGT-85.
- AABGT-72: Corrected an issue where the application would not expose the microSD card when the card was inserted into the I/O board and the card had no files on the filesystem
- AABGT-79: When connected to the Avnet IoTConnect cloud solution and the device enters, then exist low-power mode, the IoTConnect state machine has to start from the beginning sequence.
Release type: Feature Release
Date: July 18, 2023
- AABGT-60: Added USB PMSC feature to expose the RASynBoard's microSD card to the development PC when connected to the RASynBoard USB-C connector on the core board. This allows the developer to modify the config.ini file without removing the microSD card from the RASynBoard. Additionally audio (.wav) and imu (.csv) recording files can be copied from the microSD card without removing the microSD card from the RASynBoard.
- AABGT-50: Added new config.ini entry [BLE Mode] --> BLE_Enabled=x to control the application broadcasting inference results over BLE. The feature defaults to off (don't send inference results over BLE)
- AABGT-63: When [Low Power Mode] --> Power_Mode=1 is enabled, the application will automatically enter low power mode after startup. Note when this feature is enabled the USB PMSC feature is not avaliable since the device stays in low power mode until an inference result is detected.
- AABGT-70: Corrected an issue where audio and imu capture files on the microSD card would be overwritten if the application was restarted.
- AABGT-72: If the microSD card does not have any files the USB PMSC feature fails.
Release type: Feature Release
- Supports capturing 6-axis IMU data to the debug uart or to a file
- IMU data capture follows
Recording_Period=xxconfiguration item - Added Edge Impulse Data Uploader instructions
- Added Edge Impulse Data Forwarder instructions
- When the core board boots without a core board, send uart debug to USB-C on core board
- When running the circular motion model
mode=3the application hangs when the user presses the user button to record imu data.- To capture IMU data, select
mode=1and add a new lineButton_shift=imuat the end of the[Function_1]block
- To capture IMU data, select
- When in low-power mode
Power_Mode=1the application does not start-up in low-power mode- The application will enter low-power mode after the first inference result is detected
Release type: Feature Release
Initial release
- Supports selecting which ML model to run using the config.ini file for configuration options
- Supports audio data for ML models
- Supports IMU data for ML models
- Record training data with short button press < 400 ms
- Records audio data by default
- Records IMU data with config.ini option:
Button_shift=imu
- Write ndp120 images to SPI Flash with long button press > 3 sec
- Enters low-power mode when
Mode=1ormode=2and the "down down" command is detected - Supports always return to low-power mode when
Power_Mode=1in the config.ini file - Supports user defined recording period by setting
Recording_Period=xxin the config.ini file
Audio model detects 5-keywords using the RASynBoard's build in digital microphone
- "ok syntiant"
- "up"
- "down"
- "back next"
Audio model detects 5-keywords using the RASynBoard's optional 2-microphone hardware
- "ok syntiant"
- "up"
- "down"
- "back next"
IMU data model detects when the RASynBoard is moved in a circle while laying flat on a table
- When running the circular motion model
mode=3the application hangs when the user presses the user button to record imu data.- To capture IMU data, select
mode=1and add a new lineButton_shift=imuat the end of the[Function_1]block
- To capture IMU data, select
- When in low-power mode
Power_Mode=1the application does not start-up in low-power mode- The application will enter low-power mode after the first inference result is detected