Edge Application SDK
AI Model and PPL Deploy
Functional Specifications

TOC

1. Change history
2. Terms/Abbreviations
3. Reference materials
4. Expected use case
5. Functional overview/Algorithm
6. User interface specifications (Deploy AI model)
7. User interface specifications (Cancel AI model deployment state)
8. User interface specifications (Delete deployment configuration)
9. User interface specifications (Deploy "PPL")
10. Target performances/Impact on performances
11. Assumption/Restriction
12. Remarks
13. Unconfirmed items

1. Change history

Date	What/Why
2023/01/30	Initial draft.
2023/05/26	Fixed the notation of tool names and parentheses. Some environments do not render AsciiDoc’s Mermaid diagrams, so modified to rendered image references. Added alternate text to images.

2. Terms/Abbreviations

Terms/Abbreviations	Meaning
Deployment configuration	Used to deploy AI models. Specify the AI model to deploy, etc. and register it in "Console for AITRIOS". Run a deployment by specifying a registered deployment configuration. Note that it is separate from the `configuration.json` to configure notebook runtime settings.
"PPL"	A module that processes the output of the AI model (Output Tensor) of edge AI devices

3. Reference materials

Reference/Related documents
- API Reference
  - https://developer.aitrios.sony-semicon.com/development-guides/reference/api-references/
- "Console User Manual"
  - https://developer.aitrios.sony-semicon.com/en/documents/console-user-manual

4. Expected use case

Deploy an AI model imported into "Console for AITRIOS" to edge AI devices
Deploy a "PPL" imported into "Console for AITRIOS" to edge AI devices
Check AI model or "PPL" deployment state

5. Functional overview/Algorithm

Functional overview

Users can use "Console Access Library" in the SDK’s Dev Container (Local PC or Codespaces)
- Users can do the following through the "Console Access Library":
  - Get device information for deployment
  - Register a new deployment configuration to deploy AI model
  - Get a list of registered deployment configurations
  - Delete a registered deployment configuration
  - Deploy an AI model to edge AI devices
  - Deploy a "PPL" to edge AI devices

Note	To run inference after deploying an AI model, set the model ID using command parameter of the "Console for AITRIOS". These are beyond the scope of this book. See "Console User Manual" for details.

Flow

flowchart TD;
    %% definition
    classDef object fill:#FFE699, stroke:#FFD700
    classDef external_service fill:#BFBFBF, stroke:#6b8e23, stroke-dasharray: 10 2
    style legend fill:#FFFFFF,stroke:#000000

    %% impl
    subgraph legend["Legend"]
        process(Processing/User behavior)
    end

Loading

Deploy AI model

Flow

flowchart TD
    %% definition
    classDef object fill:#FFE699, stroke:#FFD700

    start((Start))
    id2(Run the notebook for system client authentication)
    id3(Run the notebook to get imported AI model list)
    id4(Run the notebook to get a list of edge AI devices)
    id5(Run the notebook to get deployment configuration list)
    id6(Create and edit the configuration file to deploy AI model)
    id7(Run the notebook to deploy AI model)
    finish(((Finish)))

    %% impl
    start --> id2
    id2 --> id3
    id3 --> id4
    id4 --> id5
    id5 --> id6
    id6 --> id7
    id7 --> finish

Loading

Flow details
1. Run the notebook for system client authentication
2. Run the notebook to get imported AI model list
  - Run the notebook to get a list of AI models that have been imported into "Console for AITRIOS", and get settings in the configuration file, model_id.
    
    See configuration.json for details.
3. Run the notebook to get a list of edge AI devices
  - Run the notebook to get a list of edge AI devices registered in "Console for AITRIOS", and get settings in the configuration file, device_id, model_id, and model_version.
    
    See configuration.json for details.
4. Run the notebook to get deployment configuration list
  - Get deployment configuration to deploy AI model
    
    Run the notebook to get a list of deployment configurations registered in "Console for AITRIOS", and get settings in the configuration file, config_id.
    See configuration.json for details.
5. Create and edit the configuration file to deploy AI model
  - Create and edit the configuration file configuration.json to configure notebook runtime settings
6. Run the notebook to deploy AI model

Cancel AI model deployment state

Flow

Note

"Cancel AI model deployment state" is to reset state on the database.
Use when a edge AI device stops responding after deploying AI model, leaving deployment state "running" on database.
You can’t rerun deployment in this state and must cancel.
(Do not rerun the notebook to deploy AI model in this state.)
You can’t cancel deployment to edge AI devices.
You can’t recover that edge AI device stops responding by SDK.
Restart or reset by other means.

flowchart TD
    %% definition
    classDef object fill:#FFE699, stroke:#FFD700

    start((Start))
    id1("Run the notebook to deploy AI model <br> ※Omit detailed flow")
    id2(Check AI model deployment state)
    id3(Create and edit the configuration file to cancel AI model deployment state)
    id4(Run the notebook to cancel AI model deployment state)
    finish(((Finish)))

    %% impl
    start --> id1
    id1 --> id2
    id2 --> id3
    id3 --> id4
    id4 --> finish

Loading

Flow details
1. Run the notebook to deploy AI model
  - See flow for details
2. Check AI model deployment state
  - Run the notebook to deploy AI model and check the deployment results
3. Create and edit the configuration file to cancel AI model deployment state
  - Create and edit the configuration file configuration.json to configure notebook runtime settings
4. Run the notebook to cancel AI model deployment state

Delete deployment configuration

Flow

flowchart TD
    %% definition
    classDef object fill:#FFE699, stroke:#FFD700

    start((Start))
    id1(Run the notebook for system client authentication)
    id2(Run the notebook to get deployment configuration list)
    id3(Create and edit the configuration file for running the notebook to delete deployment configuration)
    id4(Run the notebook to delete deployment configuration)
    finish(((Finish)))

    %% impl
    start --> id1
    id1 --> id2
    id2 --> id3
    id3 --> id4
    id4 --> finish

Loading

Flow details
1. Run the notebook for system client authentication
2. Run the notebook to get deployment configuration list
  - Run the notebook to get a list of deployment configurations registered in "Console for AITRIOS", and get settings in the configuration file, config_id.
3. Create and edit the configuration file for running the notebook to delete deployment configuration
  - Create and edit the configuration file configuration.json to configure notebook runtime settings
4. Run the notebook to delete deployment configuration
  - Run the notebook to delete deployment configuration specified in the configuration file from "Console for AITRIOS"

Deploy "PPL"

Flow

flowchart TD
    %% definition
    classDef object fill:#FFE699, stroke:#FFD700

    start((Start))
    id1(Run the notebook for system client authentication)
    id2(Run the notebook to get imported PPL list)
    id3(Run the notebook to get a list of edge AI devices)
    id4(Create and edit the configuration file for running the notebook to deploy PPL)
    id5(Run the notebook to deploy PPL)
    finish(((Finish)))

    %% impl
    start --> id1
    id1 --> id2
    id2 --> id3
    id3 --> id4
    id4 --> id5
    id5 --> finish

Loading

Flow details
1. Run the notebook for system client authentication
2. Run the notebook to get imported "PPL" list
  - Run the notebook to get a list of "PPL" that have been imported into "Console for AITRIOS", and get settings in the configuration file, app_name and version_number.
    
    See configuration.json for details.
3. Run the notebook to get a list of edge AI devices
  - Run the notebook to get a list of edge AI devices registered in "Console for AITRIOS", and get settings in the configuration file, device_id.
    
    See configuration.json for details.
4. Create and edit the configuration file for running the notebook to deploy "PPL"
  - Create and edit the configuration file configuration.json to configure notebook runtime settings
5. Run the notebook to deploy "PPL"

Sequence

Deploy AI model

%%{init:{'themeCSS':'text.actor {font-size:18px !important;} .messageText {font-size:18px !important;} .labelText {font-size:18px !important;} .loopText {font-size:18px !important;} .noteText {font-size:18px !important;}'}}%%
sequenceDiagram
  participant User
  participant Dev Container
  participant Console as Console <br> for AITRIOS

  User->>Dev Container: Run the notebook <br> for system client authentication
  opt Run arbitrarily <br> to get information needed for deployment
    User->>Dev Container: Run the notebook <br> to get imported AI model list
    Dev Container->>Console: Run the API <br> to get AI model list
    Console-->>Dev Container: Response
    Dev Container-->>User: AI model list
    User->>Dev Container: Run the notebook <br> to get a list of edge AI devices
    Dev Container->>Console: Run the API <br> to get a list edge AI devices
    Console-->>Dev Container: Response
    Dev Container-->>User: A list of edge AI devices
    User->>Dev Container: Run the notebook <br> to get deployment <br> configuration list
    Dev Container->>Console: Run the API <br> to get deployment <br> configuration list
    Console-->>Dev Container: Response
    Dev Container-->>User: Deployment configuration list
  end

  User->>Dev Container: Create and edit <br> the configuration file <br> to deploy AI model
  User->>Dev Container: Run the notebook <br> to deploy AI model
  opt To register a new deployment configuration <br> (Specify true/false in the configuration file)
    Dev Container->>Console: Run the API <br> to register deployment configuration
    Console-->>Dev Container: Response
    Dev Container-->>User: Results
  end
  Dev Container->>Console: Run the API to deploy
  Console-->>Dev Container: Response
  Dev Container-->>User: Results
  Dev Container->>Console: Run the API <br> to get deployment results
  Console-->>Dev Container: Response
  Dev Container-->>User: Results

  Note over User, Console: The API to deploy AI model is asynchronous, <br> the response is returned before the deployment is complete. <br> Run the cell to get deployment results, <br> to check the success or failure of the deployment

  opt Run arbitrarily multiple times <br> to check deployment state
    User->>Dev Container: Run the notebook <br> to deploy AI mode <br> (Run the cell to get <br> deployment results only)
    Dev Container->>Console: Run the API <br> to get deployment results
    Console-->>Dev Container: Response
    Dev Container-->>User: Results
  end

Loading

Cancel AI model deployment state

%%{init:{'themeCSS':'text.actor {font-size:18px !important;} .messageText {font-size:18px !important;} .labelText {font-size:18px !important;} .loopText {font-size:18px !important;} .noteText {font-size:18px !important;}'}}%%
sequenceDiagram
  participant User
  participant Dev Container
  participant Console as Console <br> for AITRIOS

  Note over User, Dev Container: Run when you see <br> an edge AI device <br> stops responding <br> after deploying an AI model

  User->>Dev Container: Run the notebook <br> to deploy AI mode <br> (Run the cell to get <br> deployment results only)
  Dev Container->>Console: Run the API <br> to get deployment results
  Console-->>Dev Container: Response
  Dev Container-->>User: Results

  User->>Dev Container: Create and edit <br> the configuration file <br> to cancel AI model <br> deployment state
  User->>Dev Container: Run the notebook <br> to cancel AI model <br> deployment state

  Dev Container->>Console: Run the API <br> to cancel AI model <br> deployment state
  Console-->>Dev Container: Response
  Dev Container-->>User: Results

Loading

Delete deployment configuration

%%{init:{'themeCSS':'text.actor {font-size:18px !important;} .messageText {font-size:18px !important;} .labelText {font-size:18px !important;} .loopText {font-size:18px !important;} .noteText {font-size:18px !important;}'}}%%
sequenceDiagram
  participant User
  participant Dev Container
  participant Console as Console <br> for AITRIOS

  User->>Dev Container: Run the notebook <br> for system client authentication
  opt Run arbitrarily <br> to get information you need
    User->>Dev Container: Run the notebook <br> to get deployment <br> configuration list
    Dev Container->>Console: Run the API <br> to get deployment <br> configuration list
    Console-->>Dev Container: Response
    Dev Container-->>User: Deployment configuration list
  end
  User->>Dev Container: Create and edit <br> the configuration file <br> to delete deployment <br> configuration
  User->>Dev Container: Run the notebook <br> to delete deployment <br> configuration

  Dev Container->>Console: Run the API <br> to delete deployment <br> configuration
  Console-->>Dev Container: Response
  Dev Container-->>User: Results

Loading

Deploy "PPL"

%%{init:{'themeCSS':'text.actor {font-size:18px !important;} .messageText {font-size:18px !important;} .labelText {font-size:18px !important;} .loopText {font-size:18px !important;} .noteText {font-size:18px !important;}'}}%%
sequenceDiagram
  participant User
  participant Dev Container
  participant Console as Console <br> for AITRIOS

  User->>Dev Container: Run the notebook <br> for system client authentication

  opt Run arbitrarily <br> to get information needed for deployment
    User->>Dev Container:Run the notebook <br> to get imported PPL list
    Dev Container->>Console:Run the API <br> to get PPL list
    Console-->>Dev Container: Response
    Dev Container-->>User: PPL list
    User->>Dev Container: Run the notebook <br> to get a list of edge AI devices
    Dev Container->>Console:Run the API <br> to get a list edge AI devices
    Console-->>Dev Container: Response
    Dev Container-->>User: A list of edge AI devices
  end

  User->>Dev Container: Create and edit <br> the configuration file <br> to deploy PPL
  User->>Dev Container: Run the notebook <br> to deploy PPL
  Dev Container->>Console: Run the API <br> to deploy PPL
  Console-->>Dev Container: Response
  Dev Container-->>User: Results
  Dev Container->>Console: Run the API <br> to get deployment results
  Console-->>Dev Container: Response
  Dev Container-->>User: Results
      Note over User, Console: The API to deploy PPL is asynchronous, <br> the response is returned before the deployment is complete. <br> Run the cell to get deployment results, <br> to check the success or failure of the deployment

  opt Run arbitrarily multiple times <br> to check deployment state
    User->>Dev Container: Run the notebook <br> to deploy PPL <br> (Run the cell to get <br> deployment results only)
    Dev Container->>Console: Run the API <br> to get deployment results
    Console-->>Dev Container: Response
    Dev Container-->>User: Results
  end

Loading

6. User interface specifications (Deploy AI model)

Prerequisite

You have registered as a user through "Portal for AITRIOS" and participated in the AITRIOS project
You have uploaded an AI model to the "Console for AITRIOS"

How to start each function

Launch the SDK environment and preview the README.md in the top directory
Jump to the README.md in the tutorials directory from the hyperlink in the SDK environment top directory
Jump to the README.md in the 3_prepare_model directory from the hyperlink in the README.md in the tutorials directory
Jump to the README.md in the develop_on_sdk directory from the hyperlink in the README.md in the 3_prepare_model directory
Jump to the README.md in the 4_deploy_to_device directory from the hyperlink in the README.md in the develop_on_sdk directory
Jump to the README.md in the deploy_to_device directory from the hyperlink in the README.md in the 4_deploy_to_device directory
Jump to each feature from each file in the deploy_to_device directory

Run the notebook for system client authentication

Jump to the README.md in the set_up_console_client directory from the hyperlink in the README.md in the deploy_to_device directory
Open the notebook for system client authentication, *.ipynb, in the set_up_console_client directory, and run the python scripts in it

Run the notebook to get imported AI model list

Jump to the README.md in the get_model_list directory from the hyperlink in the README.md in the deploy_to_device directory
Open the notebook to get AI model list, *.ipynb, in the get_model_list directory, and run the python scripts in it
- If successful, information about the AI models imported into "Console for AITRIOS", such as model ID, version, etc., is displayed in the notebook

Run the notebook to get a list of edge AI devices

Jump to the README.md in the get_device_list directory from the hyperlink in the README.md in the deploy_to_device directory
Open the notebook to get a list of edge AI devices, *.ipynb, in the get_device_list directory, and run the python scripts in it
- If successful, information about the edge AI devices registered in "Console for AITRIOS", such as device ID, deployed model ID, etc., is displayed in the notebook

Run the notebook to get deployment configuration list

Jump to the README.md in the get_deploy_config directory from the hyperlink in the README.md in the deploy_to_device directory
Open the notebook to get deployment configuration list, *.ipynb, in the get_deploy_config directory, and run the python scripts in it
- If successful, information about the deployment configurations registered in "Console for AITRIOS", such as config ID, etc., is displayed in the notebook

Create and edit the configuration file to deploy AI model

Note	All parameters are required, unless otherwise indicated.

Note	All values are case sensitive, unless otherwise indicated.

Note	The parameters passed to the "Console Access Library" API are as specified in the "Console Access Library" API.

Create and edit the configuration file, configuration.json, in the deploy_to_device directory.

Configuration		Meaning	Range	Remarks
`should_create_deploy_config`		Whether to register new deployment configuration	true or false true:New registration false:Use registered	Don’t abbreviate
`config_id`		ID of the deployment configuration Specify any character string for new registration If using registered, specify its ID	String Details follow the "Console Access Library" API specification.	Don’t abbreviate Used for the following "Console Access Library" API. `deployment.deployment.Deployment.create_deploy_configuration` `deployment.deployment.Deployment.deploy_by_configuration`
`create_config`	`comment`	Description of the newly registered deployment configuration	String Details follow the "Console Access Library" API specification.	Optional Use to register a new deployment configuration. Used for the following "Console Access Library" API. `deployment.deployment.Deployment.create_deploy_configuration`
	`model_id`	ID of the AI model to deploy Specify the ID of an imported AI model	String Details follow the "Console Access Library" API specification.	Optional. But don’t abbreviate this to register a new deployment configuration. Use to register a new deployment configuration. Used for the following "Console Access Library" API. `deployment.deployment.Deployment.create_deploy_configuration`
	`model_version_number`	Version of the AI model to deploy Specify the version of an imported AI model	String Details follow the "Console Access Library" API specification.	Optional Use to register a new deployment configuration. Used for the following "Console Access Library" API. `deployment.deployment.Deployment.create_deploy_configuration`
`device_ids`		ID of the edge AI devices to deploy AI model	List of strings	Don’t abbreviate Used for the following "Console Access Library" API. `deployment.deployment.Deployment.deploy_by_configuration`
`replace_model_id`		ID of the AI model to be replaced Specify the ID of the AI model to replace (overwrite) among the models deployed on the device	String Details follow the "Console Access Library" API specification.	Optional Optional if you don’t replace the AI model. (If not specified when the number of models deployed on the edge AI device has reached the limit, an error occurs.) Used for the following "Console Access Library" API. `deployment.deployment.Deployment.deploy_by_configuration`
`comment`		Deployment description	String Details follow the "Console Access Library" API specification.	Optional Used for the following "Console Access Library" API. `deployment.deployment.Deployment.deploy_by_configuration`

Run the notebook to deploy AI model

Open the notebook, deploy_to_device.ipynb, in the deploy_to_device directory, and run the python scripts in it
- The scripts do the following:
  - Checks that configuration.json exists in the deploy_to_device directory
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Checks the contents of configuration.json
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Checks the contents of configuration.json for should_create_deploy_config
    
    If true, run the API to register deployment configuration
    
    If the deployment configuration is successfully registered, deploy_to_device.ipynb displays a successful message
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Run the API to deploy AI model
    
    If API execution is successful, deploy_to_device.ipynb displays a successful message
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Run the API to get AI model deployment results
    
    If results are gotten successfully, deploy_to_device.ipynb displays a successful message and deployment results
    
    If an error occurs, the error description is displayed and running is interrupted.
    
    See "Cloud SDK Console Access Library (Python) Functional Specifications" for details on errors and response times

7. User interface specifications (Cancel AI model deployment state)

Note	Use when a edge AI device stops responding after an AI model deployment and the deployment state on the database remains "running". Operation is not guaranteed when this function is executed under normal conditions.

Prerequisite

You have registered as a user through "Portal for AITRIOS" and participated in the AITRIOS project
After deploying an AI model, check the deployment state to determine whether to cancel the deployment state

How to start each function

Launch the SDK environment and preview the README.md in the top directory
Jump to the README.md in the tutorials directory from the hyperlink in the SDK environment top directory
Jump to the README.md in the 3_prepare_model directory from the hyperlink in the README.md in the tutorials directory
Jump to the README.md in the develop_on_sdk directory from the hyperlink in the README.md in the 3_prepare_model directory
Jump to the README.md in the 4_deploy_to_device directory from the hyperlink in the README.md in the develop_on_sdk directory
Jump to the README.md in the cancel_deploy_state directory from the hyperlink in the README.md in the 4_deploy_to_device directory
Jump to each feature from each file in the cancel_deploy_state directory

Create and edit the configuration file to cancel AI model deployment state

Note	All parameters are required, unless otherwise indicated.

Note	The parameters passed to the "Console Access Library" API are as specified in the "Console Access Library" API.

Create and edit the configuration file, configuration.json, in the cancel_deploy_state directory.

Configuration	Meaning	Range	Remarks
`device_id`	ID of the edge AI device to cancel deployment state	String Details follow the "Console Access Library" API specification.	Don’t abbreviate Used for the following "Console Access Library" API. `deployment.deployment.Deployment.cancel_deployment`
`deploy_id`	Deployment ID to cancel deployment state	String Details follow the "Console Access Library" API specification.	Don’t abbreviate Used for the following "Console Access Library" API. `deployment.deployment.Deployment.cancel_deployment`

Note	After running the notebook to deploy AI model, deployment results and state are displayed. Get settings in the configuration file, `device_id` and `deploy_id` from them.

Run the notebook to cancel AI model deployment state

Open the notebook, cancel_deploy_state.ipynb, in the cancel_deploy_state directory, and run the python scripts in it
- The scripts do the following:
  - Checks that configuration.json exists in the cancel_deploy_state directory
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Checks the contents of configuration.json
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Run the API to cancel AI model deployment state
    
    If API execution is successful, cancel_deploy_state.ipynb displays a successful message
    
    If an error occurs, the error description is displayed and running is interrupted.

Note	When the API is executed, the deployment state on the "Console" transitions from "Running" to "Canceled".

8. User interface specifications (Delete deployment configuration)

Prerequisite

You have registered as a user through "Portal for AITRIOS" and participated in the AITRIOS project
You have registered a deployment configuration in the "Console for AITRIOS"

How to start each function

Launch the SDK environment and preview the README.md in the top directory
Jump to the README.md in the tutorials directory from the hyperlink in the SDK environment top directory
Jump to the README.md in the 3_prepare_model directory from the hyperlink in the README.md in the tutorials directory
Jump to the README.md in the develop_on_sdk directory from the hyperlink in the README.md in the 3_prepare_model directory
Jump to the README.md in the 4_deploy_to_device directory from the hyperlink in the README.md in the develop_on_sdk directory
Jump to the README.md in the delete_deploy_config directory from the hyperlink in the README.md in the 4_deploy_to_device directory
Jump to each feature from each file in the delete_deploy_config directory

Run the notebook for system client authentication

Jump to the README.md in the set_up_console_client directory from the hyperlink in the README.md in the delete_deploy_config directory
Open the notebook for system client authentication, *.ipynb, in the set_up_console_client directory, and run the python scripts in it

Run the notebook to get deployment configuration list

Jump to the README.md in the get_deploy_config directory from the hyperlink in the README.md in the delete_deploy_config directory
Open the notebook to get deployment configuration list, *.ipynb, in the get_deploy_config directory, and run the python scripts in it
- If successful, information about the deployment configurations registered in "Console for AITRIOS", such as config ID, etc., is displayed in the notebook

Create and edit the configuration file for running the notebook to delete deployment configuration

Note	All parameters are required, unless otherwise indicated.

Note	The parameters passed to the "Console Access Library" API are as specified in the "Console Access Library" API.

Create and edit the configuration file, configuration.json, in the delete_deploy_config directory.

Configuration	Meaning	Range	Remarks
`config_id`	ID of the deployment configuration to delete	String Details follow the "Console Access Library" API specification.	Don’t abbreviate Used for the following "Console Access Library" API. `deployment.deployment.Deployment.delete_deploy_configuration`

Run the notebook to delete deployment configuration

Open the notebook, delete_deploy_config.ipynb, in the delete_deploy_config directory, and run the python scripts in it
- The scripts do the following:
  - Checks that configuration.json exists in the delete_deploy_config directory
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Checks the contents of configuration.json
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Runs the API to delete deployment configuration
    
    If deletion is successful, delete_deploy_config.ipynb displays a successful message
- If an error occurs, the error description is displayed in the delete_deploy_config.ipynb and running is interrupted.
  - See "Cloud SDK Console Access Library (Python) Functional Specifications" for details on errors and response times

9. User interface specifications (Deploy "PPL")

Prerequisite

You have registered as a user through "Portal for AITRIOS" and participated in the AITRIOS project
You have uploaded a "PPL" to the "Console for AITRIOS"

How to start each function

Launch the SDK environment and preview the README.md in the top directory
Jump to the README.md in the tutorials directory from the hyperlink in the SDK environment top directory
Jump to the 4_prepare_application directory from the hyperlink in the README.md in the tutorials directory
Jump to the README.md in the 3_deploy_to_device directory from the hyperlink in the README.md in the 4_prepare_application directory
Jump to each feature from each file in the 3_deploy_to_device directory

Run the notebook for system client authentication

Jump to the README.md in the set_up_console_client directory from the hyperlink in the README.md in the 3_deploy_to_device directory
Open the notebook for system client authentication, *.ipynb, in the set_up_console_client directory, and run the python scripts in it

Run the notebook to get imported "PPL" list

Jump to the README.md in the get_application_list directory from the hyperlink in the README.md in the 3_deploy_to_device directory
Open the notebook to get "PPL" information list, *.ipynb, in the get_application_list directory, and run the python scripts in it
- If successful, information about the "PPL" imported into "Console for AITRIOS", such as application name, version, etc., is displayed in the notebook

Run the notebook to get a list of edge AI devices

Jump to the README.md in the get_device_list directory from the hyperlink in the README.md in the 3_deploy_to_device directory
Open the notebook to get a list of edge AI devices, *.ipynb, in the get_device_list directory, and run the python scripts in it
- If successful, information about the edge AI devices registered in "Console for AITRIOS", such as device ID, etc., is displayed in the notebook

Create and edit the configuration file for running the notebook to deploy "PPL"

Note	All parameters are required, unless otherwise indicated.

Note	The parameters passed to the "Console Access Library" API are as specified in the "Console Access Library" API.

Create and edit the configuration file, configuration.json, in the 3_deploy_to_device directory.

Configuration	Meaning	Range	Remarks
`app_name`	Name of the "PPL" to deploy	String Details follow the "Console Access Library" API specification.	Don’t abbreviate Used for the following "Console Access Library" API. `deployment.deployment.Deployment.deploy_device_app` `deployment.deployment.Deployment.get_device_app_deploys`
`version_number`	Version of the "PPL" to deploy	String Details follow the "Console Access Library" API specification.	Don’t abbreviate Used for the following "Console Access Library" API. `deployment.deployment.Deployment.deploy_device_app` `deployment.deployment.Deployment.get_device_app_deploys`
`device_ids`	ID of edge AI device to deploy "PPL"	List of strings	Don’t abbreviate Used for the following "Console Access Library" API. `deployment.deployment.Deployment.deploy_device_app`
`comment`	"PPL" deployment description	String Details follow the "Console Access Library" API specification.	Optional Used for the following "Console Access Library" API. `deployment.deployment.Deployment.deploy_device_app`

Run the notebook to deploy "PPL"

Open the notebook, deploy_to_device.ipynb, in the 3_deploy_to_device directory, and run the python scripts in it
- The scripts do the following:
  - Checks that configuration.json exists in the 3_deploy_to_device directory
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Checks the contents of configuration.json
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Run the API to deploy "PPL"
    
    If API execution is successful, deploy_to_device.ipynb displays a successful message
    
    If an error occurs, the error description is displayed and running is interrupted.
  - Run the API to get "PPL" deployment results
    
    If results are gotten successfully, deploy_to_device.ipynb displays a successful message and deployment results
    
    If an error occurs, the error description is displayed and running is interrupted.
    
    See "Cloud SDK Console Access Library (Python) Functional Specifications" for details on errors and response times

10. Target performances/Impact on performances

Usability
- When the SDK environment is built, AI models and "PPL" can be deployed from "Console for AITRIOS" to edge AI devices without any additional installation steps
- UI response time of 1.2 seconds or less
- If processing takes more than 5 seconds, indicates that processing is in progress with successive updates
- Provides users with documentation of usage tools and version information

11. Assumption/Restriction

You can’t cancel deployment or deletion of deployment configuration halfway
If you cancel and restart notebooks, start each process from the beginning instead of resuming in the middle

12. Remarks

None

13. Unconfirmed items

None

Files

EdgeApplicationSDK_FuncSpec_ModelAndPPLDeploy.adoc

Latest commit

History

EdgeApplicationSDK_FuncSpec_ModelAndPPLDeploy.adoc

File metadata and controls

Edge Application SDK AI Model and PPL Deploy Functional Specifications

1. Change history

2. Terms/Abbreviations

3. Reference materials

4. Expected use case

5. Functional overview/Algorithm

Functional overview

Flow

Deploy AI model

Cancel AI model deployment state

Delete deployment configuration

Deploy "PPL"

Sequence

Deploy AI model

Cancel AI model deployment state

Delete deployment configuration

Deploy "PPL"

6. User interface specifications (Deploy AI model)

Prerequisite

How to start each function

Run the notebook for system client authentication

Run the notebook to get imported AI model list

Run the notebook to get a list of edge AI devices

Run the notebook to get deployment configuration list

Create and edit the configuration file to deploy AI model

Run the notebook to deploy AI model

7. User interface specifications (Cancel AI model deployment state)

Prerequisite

How to start each function

Create and edit the configuration file to cancel AI model deployment state

Run the notebook to cancel AI model deployment state

8. User interface specifications (Delete deployment configuration)

Prerequisite

How to start each function

Run the notebook for system client authentication

Run the notebook to get deployment configuration list

Create and edit the configuration file for running the notebook to delete deployment configuration

Run the notebook to delete deployment configuration

9. User interface specifications (Deploy "PPL")

Prerequisite

How to start each function

Run the notebook for system client authentication

Run the notebook to get imported "PPL" list

Run the notebook to get a list of edge AI devices

Create and edit the configuration file for running the notebook to deploy "PPL"

Run the notebook to deploy "PPL"

10. Target performances/Impact on performances

11. Assumption/Restriction

12. Remarks

13. Unconfirmed items

Edge Application SDK
AI Model and PPL Deploy
Functional Specifications