Amazon Alexa通过云控制Bluetooth Mesh设备

Amazon Alexa通过云控制Bluetooth Mesh设备

The article is released under license CC BY-NC-ND 4.0

The IoT Boot Camp series course is a project initiated by the TorchIoTBootCamp team and is a joint project involving a lot of contributors working in the IoT field. It’s designed to deliver IoT knowledge efficiently that enable the connection of things, people and information.

Click here for more video courses
Click here for more courses

1. Project Background:

As the development of the smart home market, the need for controlling smart devices using human voice grows rapidly. The leading companies all over the world introduced many different solutions in the area of smart home to satisfy their customers’ requirements. Among those solutions, the most popular pattern is to control smart devices via a central smart speaker, which listens to the user’s voice command and responds with actions, such as turning on/off smart lights, adjusting the room temperature through sending remote signals to the AC, and even cooking a nice meal by operating an intelligent robot.
However, one thing often happens that breaks this good image is that the customers often find the smart speaker does not support their smart devices. That happens mainly due to one big reason—the communicating protocols used by the smart speaker and the endpoint device are not same.
Currently, there are three protocols mostly used in the smart home industry: Bluetooth mesh, Zigbee, and Z-Wave. In Chinese and the U.S. market, Bluetooth mesh and Zigbee are the mainly used protocols due to their many advantages, such as low power consumption, extendible network size, and strong security.
In the US, the Zigbee is selected by Amazon as the communicating protocol between its smart speaker brand —Amazon Echo – and the various smart devices in its ecosystem.
While in China, the Bluetooth mesh is the primary protocol that most smart devices company are using. Therefore, the Amazon Echo is not able to control those devices due to the incompatible communicating protocol.
In this project, a solution is proposed to solve this problem.
A Bluetooth mesh network is going to be controlled by an Amazon Echo Plus 2 via cloud-to-cloud approach. The user can turn on and turn off a Bluetooth mesh light by giving a voice command to the Echo Plus.

1.1. Project Preparation:

• Install the Alexa App in you iOS device. Note that the Alexa App is only available for some special regions, for e,g. US, you need to Login the AppStore with a special Apple ID. And below is an Apple ID hosted by APAC regional apps team, you can feel free to use it.

  • UserName: xxxxxxxxxxxxx_apac@outlook.com
  • Password: xxxxxxxxxxxxx

• Register the Amazon with your own account or use the available account below. For this demonstration, we have registered the account as below.

  • UserName: xxxxxxxxxxxxx_apac@outlook.com
  • Password: xxxxxxxxxxxxx

• Also you can / should Sign-in the Alexa App with the same account as above.

• If Amazon requests for account verification with email, please sign in the outlook with the account below.

  • UserName: xxxxxxxxxxxxx_apac@outlook.com
  • Password: xxxxxxxxxxxxx

• Register AWS in https://portal.aws.amazon.com/billing/signup#/start. The AWS is used for holding the lambda function, and credit card information is needed for registering a new account of the AWS.

  • The account information for AWS is not public.
  • UserName: yuancheng@xxxxxx
  • Password: xxxxxxxxxxxxx

2. Project Overview

2.1. Topology

Below is the block diagram of this project.

2.1.1. Amazon Echo Plus 2

Amazon Echo is a brand of smart speakers developed by Amazon. Echo devices connect to the voice-controlled intelligent personal assistant service Alexa, and the Echo Plus has a built-in Zigbee hub to easily setup and control your compatible smart home devices.

2.1.2. AWS Lambda

AWS Lambda lets you run code without provisioning or managing servers. You pay only for the compute time you consume - there is no charge when your code is not running.
With Lambda, you can run code for virtually any type of application or backend service - all with zero administration. Just upload your code and Lambda takes care of everything required to run and scale your code with high availability. You can set up your code to automatically trigger from other AWS services or call it directly from any web or mobile app.

2.1.3. Alexa Skill

The Alexa skill is the bridge that connects the users to AWS IoT could.
After Amazon Echo receiving the user’s voice commands, it will sent the voice record to AWS Alexa server(AI assistant server), where interprets the voice command to control directives. Then, the Alexa server sends the directives to lambda server, a place running the skill code. In the Lambda server, the specific action is executed according to the content of the directives. For example, if the user says “Alexa, turn on my light”, the Alexa will analysis this voice command and sending a JSON formatted file which listed a bunch of information about the light and actions to be executed to lambda, and then lambda will modify the light status from “OFF” to “ON” in the database.

2.1.4. AWS IoT Core

AWS IoT Core is a powerful platform to manage IoT devices remotely. It has a service called Thing Shadow, which is a JSON document recording the information of the devices it manages. In this project, the AWS IoT plays the role of the system database. It records every device’s status in the Thing Shadow document and could be visited by ESP32 via the integrated AWS freeRTOS SDK.
Thing Shadows: a JSON document that is used to store and retrieve current state information.

2.1.5. AWS FreeRTOS

AWS freeRTOS is running on ESP32 board. It connects the local network to the AWS cloud.
ESP32: ESP32 is the gateway in this project. When it runs, it continuously gets the Thing Shadow document from the AWS IoT Core, and then it translates the JSON directives to simple character strings. Through UART communication, it passes the commands to the provisioner in the Bluetooth Mesh network, and the provisioner will give the specific directives to appointed devices.

2.1.6. Bluetooth mesh network:

In this project, a simple Bluetooth network is created that includes two mesh node and of course a provisioner.
The provisioner in the network to provision new devices and receive commands from ESP32 via UART, and also parse the commands from ESP32 and transfer to the mesh command and transmit it to the mesh network for controlling other nodes.
There are total two Bluetooth mesh nodes in the network, a light node and a switch node. Once the user presses the button on the switch node, the light node would receive a BLE mesh message and turn on/off the embedded LED on it.

Also, the switch node is capable to change the online shadow document. When the user give it a long press on left button, the switch node will send a message to provisioner to update the online shadow document.

2.2. Technical details:

In this section, the technical details below will be introduced.

• Alexa Smart Home Skill
• Lambda function
• ESP32 freeRTOS application
• Bluetooth Mesh provisioner

sample codes, including lambda function, Alexa smart home skill, ESP32 freeRTOS application, and the codes of the Bluetooth Mesh provisioner (based on EFR32BG13) will also be explained.

However, because of the limited length of this article, if the reader wants to replicate this project, please read the section of Replication the project for the step by step guidance, and access the GitHub page: Alexa Control Bluetooth Mesh Devices and find the mentioned packages accordingly.

2.2.1. Alexa Smart Home Skill

Alexa smart home skill interface is designed for controlling smart home devices using Amazon Echo series smart speaker by Amazon. In this project, an Echo plus 2 was utilized to receive the voice command and transmit the voice command to the Alexa server.
The skill is held by a lambda function, which means the code is running on the Amazon lambda server. When a user gives the smart speaker a voice command, the command will be firstly analyzed on Alexa server. Then, a JSON format directive will be sent to lambda. There are many different directives. The most important one is the discovery directive, which indicates the speaker to find any available devices and report back to the Alexa server.
In this project, the response is also generated by the same lambda function. After receiving the discovery directive, the lambda function will directly send a response message back. Therefore, the device information needs to be programmed in the codes.

if namespace == 'Alexa.Discovery':  
    if name == 'Discover':  
        adr = AlexaResponse(namespace='Alexa.Discovery', name='Discover.Response')  
        #create capability part for the response  
        adr = create_discover_response(adr)               
        return send_response(adr.get())   

The “AlexaResponse()” function is used to generate the response — a JSON document. It uses **kwags parameter to receive whatever the user appointed attributes. Among those attributes, the “namespace” and “name” is necessary to add. Alexa server will use this information to identify what the response is and give a feedback to the user.

#create the discover response   
#first create the capabilities that the endpoints need   
#then add the endpoints to the response entity, and add the capability information to the endpoints  
def create_discover_response(response):  
    #general response  
    capability_alexa = response.create_payload_endpoint_capability()  
    #specific capabilities  
    #power controller--turn on turn off operations  
    capability_alexa_PowerController = response.create_payload_endpoint_capability(  
        interface='Alexa.PowerController',  
        supported=[{'name': 'powerState'}])  
    #create colorcontroller capability  
    capability_alexa_ColorController = response.create_payload_endpoint_capability(  
        interface='Alexa.ColorController',  
        supported=[{'name': 'color'}])  
    #create brightnessController capability  
    capability_alexa_BrightnessController = response.create_payload_endpoint_capability(  
        interface='Alexa.BrightnessController',  
        supported=[{'name': 'brightness'}])  
    #create colortemperature capability  
    capability_alexa_ColorTemperatureController = response.create_payload_endpoint_capability(  
        interface='Alexa.ColorTemperatureController',  
        supported=[{'name': 'colorTemperatureInKelvin'}])  
    #create lock controller capability  
    capability_alexa_lockcontroller = response.create_payload_endpoint_capability(  
        interface='Alexa.LockController',  
        supported=[{'name':'lockState'}]  
    )  
    capability_alexa_endpointHealth = response.create_payload_endpoint_capability(  
        interface='Alexa.EndpointHealth',  
        supported = [{'name':'connectivity'}]  
    )       

The capabilityand endpointid compose the virtual representation of a smart device. One device can simultaneously own several different capabilities. For example, a light can support PowerController, ColorController, and ColorTemperatureController capabilities to control the on/off, color, and temperature attributes.
After constructing the capabilities of the device, we also need to add other device information and the end