WEBVTT 00:00.140 --> 00:05.960 Now that you know how to make the Amazon Alexa voice assistant communicate with our PC through a web 00:05.960 --> 00:08.060 server provided by Ngrok. 00:08.300 --> 00:12.920 Let's move on to developing the interface between the Voice Assistant and Ros. 00:12.970 --> 00:13.250 Two. 00:13.790 --> 00:19.760 We are going to create the logic that interprets the commands received from the voice assistant and 00:19.760 --> 00:21.200 convert them into Ros. 00:21.200 --> 00:24.500 Two messages that will be sent to the task server. 00:25.260 --> 00:31.140 To develop this component, we will use the documentation, the libraries and the example code provided 00:31.140 --> 00:34.710 by Amazon for the Voice assistant skilled developer. 00:35.220 --> 00:41.230 Unfortunately, though, these Alexa Voice assistant APIs are not available for C plus plus. 00:41.310 --> 00:47.580 Therefore, in this lesson we are going to implement the logic only in Python, unlike what we did in 00:47.580 --> 00:48.720 the course so far. 00:48.750 --> 00:53.640 There won't be an equivalent lesson for developing these functionalities in C plus plus. 00:54.090 --> 00:58.860 However, if you are not familiar with Python or you are not comfortable with this programming language, 00:58.860 --> 00:59.730 don't worry. 00:59.760 --> 01:05.220 Most of the code that we are going to implement in this lesson will be copy pasted from the documentation 01:05.220 --> 01:09.930 examples that are available in the Amazon Developer skill console. 01:10.290 --> 01:16.230 So instead of focusing on the language syntax, I recommend you paying attention to the logic that is 01:16.230 --> 01:22.230 used in order to add new features and new functionalities to this skill and also the logic that is used 01:22.230 --> 01:25.960 to receive and to interpret these voice messages. 01:26.500 --> 01:32.250 So let's start by creating a new Python script within the Arduino bot remote package. 01:32.260 --> 01:40.600 And so by convention we need to insert the new Python scripts within the new folder called Arduino Bot 01:42.370 --> 01:43.180 Remote. 01:43.180 --> 01:50.080 And within this folder, let's start by creating a new file called Init Dot Pi. 01:50.620 --> 01:52.780 And let's leave this file empty. 01:52.900 --> 01:58.930 And now let's create another file still in the Arduino bot remote package and in the Arduino bot remote 01:58.930 --> 01:59.530 folder. 01:59.530 --> 02:05.440 And let's call this one Alexa interface dot pi. 02:05.830 --> 02:10.630 In this script, let's start by declaring the python interpreter directory. 02:10.930 --> 02:15.280 So this is in the user bin. 02:16.180 --> 02:19.990 And the environment that we want to use is the Python three. 02:20.770 --> 02:27.820 Next, we are going to create a web server using the python flask library so we can start opening the 02:27.820 --> 02:30.580 documentation of the Alexa developer. 02:30.580 --> 02:33.040 So the Alexa developer documentation. 02:33.130 --> 02:43.000 And here let's go to the section skill development process and let's click on tools to create and manage 02:43.000 --> 02:43.810 skills. 02:45.770 --> 02:50.100 And then within this one, let's click on Ask SDK is. 02:50.790 --> 02:55.700 And here we can see all the libraries that are available for different programming languages. 02:55.710 --> 03:00.060 So let's open the one that is available for Python here. 03:00.060 --> 03:05.400 Let's open the guide on how to host a skill as a web service using flask. 03:05.490 --> 03:09.690 So this one host a custom skill as a web service. 03:10.170 --> 03:12.450 So this is a very nice tutorial. 03:12.450 --> 03:16.930 And let's scroll down to this section in which flask is used. 03:16.950 --> 03:19.170 So here, flask ask. 03:19.440 --> 03:24.870 And here we can see that there are several alternative Python libraries that are available for hosting 03:24.870 --> 03:25.380 the skill. 03:25.380 --> 03:27.330 But we are going to use this one. 03:27.330 --> 03:32.940 So the flask ask package that we installed in the setup section of this course. 03:33.330 --> 03:42.030 So now we can copy all of this contents, so we can copy this example code and we can paste it within 03:42.030 --> 03:42.960 our script. 03:42.990 --> 03:45.030 So here let's paste it. 03:45.850 --> 03:50.050 And basically with this code, we are creating a web service using flask. 03:50.050 --> 03:56.710 So using this library here and also we are initializing the communication with the Alexa skill using 03:56.740 --> 04:04.510 two components from the Alexa Ask SDK library that are the skill builder and also the skill adapter. 04:04.810 --> 04:10.330 Now, in order to start the Flask web service, we need to define the main function. 04:10.330 --> 04:14.680 So here we need to define the main function. 04:19.240 --> 04:27.790 And within this function we can execute the flask web server by using app dot run. 04:27.790 --> 04:32.950 So app will contain the implementation of the Flask web server. 04:32.950 --> 04:34.600 And here we are starting it. 04:35.570 --> 04:41.610 Then we also need to register this Flask application within the Alexa skill adapter. 04:41.630 --> 04:54.860 So within this object here, so let's use the instruction skill adapter register and let's pass the 04:54.860 --> 04:59.180 application which contains the web server is the implementation. 04:59.720 --> 05:03.500 So this one so is the implementation of the Flask class. 05:03.860 --> 05:06.650 And also let's define the route of the web service. 05:07.460 --> 05:11.210 So the route with the forward slash. 05:11.990 --> 05:18.320 Now with this, we have created our web service and also we have interfaced it with the Alexa skill 05:18.320 --> 05:21.410 adapter and also with the Skill Builder class. 05:21.530 --> 05:27.710 And now we can add all these skills functionalities and so we can add all the actions that this script 05:27.710 --> 05:28.280 will take. 05:28.280 --> 05:35.540 When a specific intent is activated, we can find this information still in the Alexa documentation. 05:35.750 --> 05:39.200 And so let's go to developing your first skill. 05:39.200 --> 05:41.390 So let's go to this tutorial here. 05:42.000 --> 05:46.990 And let's go to the section in which the first intent is executed. 05:47.010 --> 05:52.170 So here let's go to the launch request handler, which is the launch intent. 05:52.170 --> 05:55.980 And this is automatically called when we start the interaction with Alexa. 05:55.980 --> 06:04.650 So when we say the phrase Alexa, activate the robot, so now we can just copy the code for the launch 06:05.130 --> 06:09.690 request handler and we can paste it within our script. 06:09.690 --> 06:17.700 So let's paste it here Before the creation of the Skill Builder, let's move all the import statement 06:18.720 --> 06:26.070 at the beginning of this script and now we can adapt the code that we just copied for our robot. 06:26.190 --> 06:32.220 So in order to define which action should the robot take when the voice assistant is activated? 06:32.220 --> 06:36.330 So basically when this launch request is activated. 06:37.200 --> 06:39.420 Let's change the speech text. 06:39.420 --> 06:42.960 So let's change the response of the assistant. 06:43.110 --> 06:44.550 So this one. 06:45.000 --> 06:51.480 So this is the response that the assistant should provide when the launch request is activated. 06:51.600 --> 06:54.420 And let's change this one to I. 06:57.180 --> 06:58.170 How can I help? 06:58.350 --> 07:05.670 So basically, this means that whenever we say the phrase Alexa activate the robot, the assistant will 07:05.670 --> 07:09.510 answer us with this question so I can help. 07:09.930 --> 07:15.900 And actually, this is all we need in order to handle the launch request when we receive it from the 07:15.900 --> 07:16.290 Amazon. 07:16.290 --> 07:16.710 Alexa. 07:17.310 --> 07:19.560 Now we need to register this class. 07:19.560 --> 07:24.450 So the launch request handler class inside the skill builder class. 07:24.450 --> 07:31.050 So inside this class here, in fact, as we can see from the comment that we have copied, we need to 07:31.050 --> 07:39.840 register your intent handler within the skill builder object and we can do so by using the skill builder 07:39.840 --> 07:46.260 and by using the function add request handler. 07:46.620 --> 07:53.490 And to this one we can pass the instance of the launch request handler that for the moment is the only 07:53.490 --> 07:55.320 handler that we have created. 07:55.680 --> 07:58.470 And so now we can save this file here. 07:58.470 --> 08:04.800 And now we can already test this functionality by activating the skill with the phrase Alexa, activate 08:04.800 --> 08:11.520 the robot, and in order for this request handler to work, we expect the Alexa assistant to return 08:11.520 --> 08:12.600 this response. 08:12.600 --> 08:15.240 So to say this phrase here, I. 08:15.240 --> 08:16.200 How can I help? 08:16.650 --> 08:18.230 We are missing one thing. 08:18.240 --> 08:25.770 So as you can see here in the skill ID, we need to set the ID of our skill that we have just created. 08:25.950 --> 08:32.130 And in order to take this ID, we need to go back into the Amazon Alexa Developer console. 08:32.340 --> 08:39.040 And here, as you can see in the main page of our skill, we have the small button copy skill ID. 08:39.370 --> 08:44.860 So let's take this one and let's paste it here within the skill ID. 08:45.340 --> 08:47.320 So let's use the double quotes. 08:48.410 --> 08:51.320 And let's paste the SQL ID in this way. 08:51.350 --> 08:57.320 Now we have completed the connection between the Alexa console and this Python script. 08:57.470 --> 09:02.630 And now we can actually run this script in order to test its functionalities. 09:02.840 --> 09:09.080 So in order to do so, let's take the Alexa console and put it on one side. 09:09.230 --> 09:17.510 Then let's open a new terminal and let's go into the workspace, then into the source folder and into 09:17.510 --> 09:20.810 the Arduino bot remote package. 09:21.440 --> 09:27.200 Then let's go into the Arduino bot remote subfolder. 09:27.200 --> 09:34.250 And so here we have created our Alexa interface.py file and let's make this file an executable with 09:34.250 --> 09:41.360 the command C h mod plus X and then the name of the script. 09:41.360 --> 09:42.860 So Alexa interface. 09:43.160 --> 09:50.520 Now if we type ls, we can see that actually the Alexa interface dot py script is an executable and 09:50.520 --> 09:51.750 so we can execute it. 09:51.750 --> 09:54.680 We can run it as a regular python script. 09:54.690 --> 09:59.580 So this is not yet a Ros two node and we can start it as a regular python script. 09:59.580 --> 10:07.290 So with the command python followed by the name of the script that is the Alex interface. 10:07.290 --> 10:08.940 So let's press enter. 10:09.210 --> 10:12.150 This is Python three since we are using Python three. 10:13.580 --> 10:17.720 And now we can see that the Flask application is online. 10:17.720 --> 10:22.430 So it is online and it is available on the port 5000 of our PC. 10:23.090 --> 10:29.870 Now, let's also start Ngrok, because this just started the web server that listens for the new messages 10:29.870 --> 10:32.420 and interprets the new messages. 10:32.870 --> 10:38.690 Now let's also start andyrock that instead will make available a web service on our PC. 10:39.350 --> 10:44.600 So let's go to the download folder in my case, and then to the Ngrok folder. 10:44.600 --> 10:50.480 And here, in order to execute Ngrok, let's execute the file called Ngrok. 10:50.900 --> 10:59.420 And then we want to open an Http server and we want to open this one on the Port 5000 or on our PC. 11:00.050 --> 11:02.060 So let's press enter. 11:02.240 --> 11:08.120 And since we open that new instance of Ngrok and since we are using the free version of this software, 11:08.120 --> 11:13.500 it provides us with a new address to access our PC from the Internet. 11:13.500 --> 11:20.820 So as you can see, this address here is different from the one that we set previously, and so we need 11:20.820 --> 11:24.090 to change it also within the Arduino bot skill. 11:24.240 --> 11:30.540 And so within the end point here, as you can see, these two address are different since we are using 11:30.540 --> 11:32.430 the Ngrok free version. 11:32.820 --> 11:38.520 So every time we need to change this one and we need again to save the endpoint. 11:38.980 --> 11:42.970 Now everything is ready, so we can go back to the test section. 11:43.670 --> 11:52.400 So here and here we can start the interaction with our web server and also here with our Flask application 11:52.400 --> 11:58.310 that will receive the messages and will interpret them and also will provide the response that the Alexa 11:58.310 --> 12:00.110 assistant will pronounce. 12:00.350 --> 12:02.390 So let's activate this skill. 12:03.680 --> 12:05.540 Activate robot. 12:08.200 --> 12:09.790 Hi, How can I help? 12:10.150 --> 12:15.040 And now we can see that the voice assistant answered with the expected phrase. 12:15.250 --> 12:18.600 So the one that we set here, I. 12:18.610 --> 12:19.450 How can I help? 12:20.580 --> 12:26.050 And so we can see that the Rock was able to receive the message. 12:26.070 --> 12:30.560 This time the request was accepted and was processed correctly. 12:30.570 --> 12:37.260 Since we have the Flask server here that was able to interpret the request and also to send a response 12:37.260 --> 12:38.340 to the server. 12:38.340 --> 12:46.170 And so here in the Alexa we can see now both the messages, so we can see the Json message that Alexa 12:46.170 --> 12:53.310 sent to our web service and here we can see the response that our web service returned to Alexa. 12:53.310 --> 12:59.160 And also here we can see the phrase that we expect that we want Alexa to pronounce. 13:00.180 --> 13:07.470 Now all that remains is to add the endling for the remaining requests inside our script in order to 13:07.470 --> 13:10.710 complete the voice interaction model with the assistant. 13:10.740 --> 13:14.910 Now all that remains is to handle all the remaining requests. 13:14.910 --> 13:21.360 So all the remaining intents that Alexa can send to our web application inside the script that we have 13:21.360 --> 13:26.460 just created in order to complete the voice interaction model with the assistant. 13:26.460 --> 13:33.060 And then also for each of these requests, we will enable an interface with Roscoe and with the task 13:33.060 --> 13:36.300 server that we created in order to move the robot.