WEBVTT

00:00:01.560 --> 00:00:08.360
To create your simple AI exercise timer, 
you’ll train a machine learning (or ML) model. 

00:00:08.360 --> 00:00:15.080
This model will recognise when you’re 
exercising and when you’re not exercising.

00:00:15.080 --> 00:00:23.720
You’ll then combine the model with some 
ready-made code for an exercise timer...

00:00:23.720 --> 00:00:28.720
before downloading it to your 
micro:bit and using it in real life.

00:00:28.720 --> 00:00:35.960
Click ‘Open in micro:bit CreateAI’ to launch the project. 

00:00:35.960 --> 00:00:40.760
The project comes with 3 samples 
of movement data for exercising

00:00:40.760 --> 00:00:45.440
and 3 samples of movement 
data for not exercising.

00:00:45.440 --> 00:00:51.000
You’ll need to add more samples by 
recording your own movement data.

00:00:51.000 --> 00:00:56.480
micro:bit CreateAI collects movement 
data samples using the accelerometer  

00:00:56.480 --> 00:01:02.240
(or movement sensor) on the micro:bit.
You will wear a micro:bit and battery  

00:01:02.240 --> 00:01:08.840
pack on your wrist or ankle, so that you can move 
freely to record your own movement data samples. 

00:01:08.840 --> 00:01:13.480
To get started, you need to set 
up the data collection micro:bit. 

00:01:13.480 --> 00:01:22.240
Connect the wrist-worn micro:bit to CreateAI. 
If your computer has Bluetooth enabled then you  

00:01:22.240 --> 00:01:28.680
will just need 1 micro:bit and a USB data lead.
If you don’t have a Bluetooth connection,  

00:01:28.680 --> 00:01:33.800
you’ll be prompted to use 2 micro:bits.
The second micro:bit will remain connected  

00:01:33.800 --> 00:01:39.640
to the USB cable and act as a radio 
link to the data collection micro:bit.

00:01:39.640 --> 00:01:42.936
Follow the instructions on screen to connect.

00:01:44.495 --> 00:01:47.320
 Once your data collection micro:bit is connected

00:01:47.320 --> 00:01:53.280
you’ll see the lines on the live graph 
change as you move your micro:bit about.


00:01:53.280 --> 00:01:58.360
You’re now ready to add your 
own movement data samples.


00:01:58.360 --> 00:02:04.440
As this project already includes some 
data samples, we suggest you just add  

00:02:04.440 --> 00:02:12.840
1 more sample for each action for now, and spend 
more time collecting and analysing data later.

00:02:12.840 --> 00:02:16.600
Decide what ‘exercising’ 
action you are going to do. 

00:02:16.600 --> 00:02:23.040
This could be running, walking briskly, 
jumping, boxing, dancing, or any other exercise. 

00:02:23.040 --> 00:02:29.920
Make sure the micro:bit is attached to 
the wrist or ankle that will be moving. 

00:02:29.920 --> 00:02:34.440
To add data to a specific action, 
select it by clicking on it. 

00:02:34.440 --> 00:02:39.640
You will get a 3 second countdown 
before a 1 second recording starts. 

00:02:39.640 --> 00:02:45.200
Click record and start moving right away 
to ensure you get a clean data sample. 

00:02:45.200 --> 00:02:49.080
A clean sample is one where you 
are moving for the entire sample,  

00:02:49.080 --> 00:02:55.440
you don’t start late or finish moving early.
Next try adding an extra data sample  

00:02:55.440 --> 00:03:00.920
to the ‘not exercising’ data set.
Select it by clicking on the action,  

00:03:00.920 --> 00:03:06.440
then stay still, or only move very 
slightly as you record the sample. 

00:03:06.440 --> 00:03:10.800
You’ll notice that the x,y,z 
lines change places depending  

00:03:10.800 --> 00:03:14.768
on the angle at which you hold your micro:bit.

00:03:14.768 --> 00:03:18.320
The project doesn’t have a lot of data right now, 

00:03:18.320 --> 00:03:24.400
but we have enough to train our own 
machine learning model using CreateAI. 

00:03:24.400 --> 00:03:30.360
So click ‘Train’ to use the 
current data to build an ML model.

00:03:30.360 --> 00:03:34.640
The tool now builds a mathematical 
model that should recognise different  

00:03:34.640 --> 00:03:39.360
actions when you move your micro:bit.
As soon as the model has been trained,  

00:03:39.360 --> 00:03:44.080
you’ll see the Testing model page.
Your data collection micro:bit can now  

00:03:44.080 --> 00:03:50.160
be used to test how well the model is working.
It should still be connected to the tool,  

00:03:50.160 --> 00:03:59.800
and you’ll see that as you move it, CreateAI 
is estimating what action you are doing. 

00:03:59.800 --> 00:04:05.240
Try out different levels of exercising or 
not exercising to see both the estimated  

00:04:05.240 --> 00:04:12.400
action and the certainty bar graph change.
The % on the certainty bar graph shows how  

00:04:12.400 --> 00:04:16.840
confident the model is that 
you are doing each action.

00:04:16.840 --> 00:04:22.240
You may notice your model is not estimating 
some actions accurately, or maybe it is  

00:04:22.240 --> 00:04:28.560
working well for one action but not the other,
so after exploring how it is currently working,  

00:04:28.560 --> 00:04:34.520
it’s a good idea to click on ‘Edit 
data samples’ and improve your model.

00:04:34.520 --> 00:04:41.320
Machine learning models usually work best with 
more data, so record some extra samples for each  

00:04:41.320 --> 00:04:48.800
of the actions, or focus on collecting more data 
for the action that was problematic in testing.

00:04:48.800 --> 00:04:54.240
You can record 1 sample at a time or 
you can record 10 samples in sequence.

00:04:54.240 --> 00:04:58.840
Clean data samples also 
help an ML model work better 

00:04:58.840 --> 00:05:05.560
so interrogate your data set and identify any 
data samples that could confuse the model. 

00:05:05.560 --> 00:05:12.080
You can delete these by pressing x.
Once you’ve added more data and checked  

00:05:12.080 --> 00:05:17.720
your data set, click Train model 
again to use your amended data set. 

00:05:17.720 --> 00:05:27.000
Then test the model again on 
the ‘Testing model’ page.

00:05:27.000 --> 00:05:30.200
Once you’re happy with how 
the ML model is behaving,  

00:05:30.200 --> 00:05:34.120
you can use it with the ready-made project code. 

00:05:34.120 --> 00:05:40.920
Click on 'Edit in MakeCode' to see the code blocks 
in a special version of Microsoft MakeCode.

00:05:40.920 --> 00:05:48.560
You can always return to CreateAI using 
the arrow in the top left of the screen. 

00:05:48.560 --> 00:05:54.720
These code blocks use the model you 
have created within an exercise timer.

00:05:54.720 --> 00:05:59.760
The code uses two variables to keep track 
of how long you've been exercising and how  

00:05:59.760 --> 00:06:04.160
long you've not been exercising.
When the program first runs it sets  

00:06:04.160 --> 00:06:09.360
these timer variables to 0.
The 'on ML start' blocks are  

00:06:09.360 --> 00:06:15.400
triggered when the ML model decides you have 
started either exercising or not exercising. 

00:06:15.400 --> 00:06:19.240
They show different icons on the 
micro:bit's LED display depending  

00:06:19.240 --> 00:06:25.920
on the action it has estimated you are doing.
The 'on ML stop' blocks are triggered when the  

00:06:25.920 --> 00:06:32.960
ML model decides you have finished an action, 
in this case exercising or not exercising. 

00:06:32.960 --> 00:06:38.240
Code inside each block clears the screen and 
adds the duration of the action that has just  

00:06:38.240 --> 00:06:43.360
finished to the variable storing 
the total times for each action.

00:06:43.360 --> 00:06:49.680
The ML model works with the code to allow you 
to view the total time spent on each action. 

00:06:49.680 --> 00:06:55.160
Press button A to see the total time you have 
been exercising and press button B to see the  

00:06:55.160 --> 00:07:00.680
total time you have been inactive.
The timer counts in milliseconds,  

00:07:00.680 --> 00:07:07.880
thousandths of a second, so the number shown 
is divided by 1000 to show a time in seconds.

00:07:07.880 --> 00:07:12.640
To make your simple AI exercise 
timer run on your micro:bit, you  

00:07:12.640 --> 00:07:19.360
just need to download this code to a micro:bit.
If you don’t have another micro:bit available,  

00:07:19.360 --> 00:07:26.440
simply replace the code currently on the data 
collection micro:bit with the project code.


00:07:26.440 --> 00:07:31.200
Now test the project out in real life. 
Do the correct icons display  

00:07:31.200 --> 00:07:36.640
when you are exercising or not?
You can test if the timer code is working  

00:07:36.640 --> 00:07:41.640
well with the model in 3 easy steps:
Press the reset button. 

00:07:41.640 --> 00:07:46.040
Exercise for 30 seconds.
Then press button A.

00:07:46.040 --> 00:07:50.080
You should see the number 30 
scroll across your display.

00:07:50.080 --> 00:07:55.240
You’re now ready to connect to CreateAI, 
collect your own data, use it to train,  

00:07:55.240 --> 00:08:00.160
test & improve a machine learning model.
And you can then combine this model with  

00:08:00.160 --> 00:08:05.642
the ready-made code and try 
it out on your own micro:bit.