You are on page 1of 8

Educator Guide

Artificial Intelligence
45 minutes

Sustainable Farming

EDUCATION.MINECRAFT.NET

1
THEME OVERVIEW
You have been sent to a developing country to help the local population farm in a
more effective way, to make sure there is not a shortage of food. The goal is to
help optimize farming and create a more sustainable food source for the local
population. This will be accomplished by collecting satellite data in a semi-
supervised machine learning AI that utilizes predictive analysis to find areas with
the right weather and terrain conditions.

A partner called SunCulture is using these kinds of AIs to help farmers in


developing countries make the most use of their land and maximize their crop
yield. For more information on their activities:
https://www.microsoft.com/en-us/ai/ai-for-earth-SunCulture

LESSON OBJECTIVES
By the end of the lesson, students will:
• Gather and prepare satellite data for a dataset
• Understand the concept and application of predictive analysis
• Use machine learning algorithms to improve crop yields and soil efficiency
in developing countries
• Create, test, and debug code
• Embrace a coding mindset

KEY VOCABULARY
Semi-Supervised Machine Learning – An approach to machine learning that
combines a small amount of labeled (supervised) data with a large amount of
unlabeled (unsupervised) data
Predictive Analysis– Using statistics from machine learning to make a prediction
about what will happen next
Data Bias – data that is incomplete or inherently flawed

ARTIFICIAL INTELLIGENCE CONCEPTS: EXPLANATION


In this lesson, the AI concepts presented have been selected to best explain the
process of development of a workflow involving AI within Minecraft. This does not
exactly reflect a real-world AI development process and is intended to be just
conceptual in nature. Below are the AI concepts used in the lesson in more detail
and can be used to reinforce or further expand a student’s knowledge.
• Machine Learning:
Machine learning is an important part of artificial intelligence. Machine
learning algorithms are a group of AIs that can automatically learn and

2
improve on their own without any direct human interaction. A machine
learning algorithm is an algorithm developed to be able to alter certain
parts of its code. One specific type of machine learning is:
o Semi-Supervised Machine Learning: These algorithms fall somewhere
in between supervised and unsupervised algorithms. Typically, in
these cases, a small amount of labelled data is used to start the
training of the AI after which a larger amount of unlabeled dataset is
given to the AI. One of the advantages of these algorithms is that they
are usually faster to develop than unsupervised as the training
process is faster.
o Using the example from before, if we give a semi-supervised machine
learning algorithm a relatively small dataset containing labelled
pictures of elephants and lions and then let the AI train on the
dataset, different patterns in the dataset will be found. We then give
the AI a larger dataset of unlabeled pictures, it can use what is has
learned about the animal’s characteristics from the smaller dataset to
identify and determine the animal within a certain percentage of
success.
o This machine learning method will be used in this lesson tor firstly
train the AI using labelled data about the historical weather patterns
in an area. After which the AI will be given real time data from
ground sensors and satellites that is not labelled.
• Predictive Analysis:
o Once a large enough labelled historical dataset has been collected
and prepared, an AI can predict, by comparing current data against
the historical data and looking for a similar pattern of data, the
outcome of certain situations. This is called predictive analysis and its
results are correct within a degree of accuracy.
o For example, if we give an AI a big enough dataset containing
weather information, such as, heat, wind speed, and cloud density, it
can predict crop yield.
• Data Bias:
o The term data bias means that the data is incomplete, or it is
inherently flawed. There are many kinds of biases in AI that occur in
different situations. A common situation where bias can occur in the
results is often due to the historical data which the AI was given or
trained on was biased to begin with. An example of this would be if
we are developing an AI to detect and count different types of animal
species from photographs. If the AI has been trained to count

3
animals, but the dataset was 90% bird photographs, the AI would
probably not count an accurate number of mammals due to lack of
mammal photographs it was originally given. Therefore, the total
count of animals would be incorrect.

LESSON ACTIVITIES
Direct Instruction (Teacher-Led; “I Do”)
Today, we are going to be learning about artificial intelligence and machine
learning. Our lesson is called “Sustainable Farming”. (slide 1)

Review the lesson objectives with students. (slide 2)

Review the important vocabulary with students. (slide 3)

We are going to be exploring a computer science concept called artificial


intelligence (AI). Artificial Intelligence is a computer or software that mimics the
human way of thought, such as learning and problem-solving. In the next couple
of lessons, we are going to investigate how AI concepts could be (or are being)
used in real life.

You have been sent to a developing country to help the local population farm in a
more productive way so that there is not a shortage of food. This is done by
collecting satellite data from a weather station and the using the data in a semi-
supervised machine learning AI that utilized predictive analysis to find areas with
the right weather and terrain conditions. This is going to help optimize farming
and create a more sustainable food source for the local population. (slide 4)

Guided Instruction (Teacher Modeling; “We Do”)


At this time, have all students log into Minecraft: Education Edition. Navigate to the
AI-Lesson: “Sustainable Farming”. (slide 5)

Activity 1: Withering Crops (slides 6-11)


Objective: Place Satellite Beacons on Gold Blocks then Travel to Satellite Station
(NO CODING IS REQUIRED FOR THIS ACTIVITY)
Explanation:
For this first activity, you will need to place the satellite beacons on the gold blocks
around the village. The satellite will be utilizing these beacons to orient itself. Once
done, get into the jeep to get to the satellite station

4
Independent Work (Teacher Support; “You Do”)
Activity 2: Out of this World (slides 12-16)
(There is coding in this activity)
Objective: Run Satellite and Code a Current Dataset

Explanation:
You will now need to run the satellite and code a current dataset from the data
that it is sending.

First, use a control satellite coding block inside of the on start coding block.
Then place a make dataset coding block, set to current dataset. Inside of that place
all of the get location satellite data coding blocks.

Now, when the code is run, the student will be able to see the satellite scanning
the village and collecting data.

Activity 3: Predictive Analysis (slides 17-21)


(There is coding in this activity)
Objective: Code the AI and Utilize Predictive Analysis
Explanation:
You will need to code the AI that is going to find the best locations. They will then
utilize predictive analysis and the historical data that the weather station already
has.

First, you will use a semi-supervised machine learning coding block inside of the on
start coding block.

5
• Inside of that block, place an input dataset coding block.
• Then place a compare dataset against coding block set to current and
historical.
• After that, place an run predictive analysis coding block.
• Finally, place a find best farming locations coding block.

Now, when the code is run, the AI will find the best farming locations.

Activity 4: Green Thumb (slides 22-27)


(THERE IS NO CODING IN THIS ACTIVITY)
Objective: Navigate to Farming Plots and Plant
Now that we know the best farming locations, we need to see if the AI was correct.
Navigate to those plots and press the button on the gold block to tell the farmer to
plant there.

LESSON CONCLUSION
Ask students about the skills that they have learned during the lesson to reinforce
the concepts learned. (slides 28-29)

1. What is a dataset?
Answer: A collection of information
2. Why is land optimization an important thing when farming?
Answer: By optimizing how we segregate land when farming we can
increase yields.
3. How can AI help developing countries increase their food production?
Answer: AI can help with the use of predictive analysis to find the most
optimal locations for farming.

These questions can also be used as a formative assessment. A printable version


of these questions is located at the end of this document.

6
EDUCATIONAL STANDARDS
UNITED STATES: CSTA
• 1B-DA-07 Use data to highlight or propose cause-and-effect relationships,
predict outcomes, or communicate an idea.
• 1B-AP-09 Create programs that use variables to store and modify data.
• 1B-IC-18 Discuss computing technologies that have changed the world, and
express how those technologies influence, and are influenced by, cultural
practices.
UNITED STATES: ISTE
• 1.3.d Students build knowledge by actively exploring real-world issues and
problems, developing ideas and theories and pursuing answers and
solutions.
• 1.5.b Students collect data or identify relevant data sets, use digital tools to
analyze them, and represent data in various ways to facilitate problem-
solving and decision-making.
• 1.6.c Students communicate complex ideas clearly and effectively by
creating or using a variety of digital objects such as visualizations, models or
simulations.

AUSTRALIAN F-10 CURRICULUM: DIGITAL TECHNOLOGIES (YEAR 3 AND


4)
• (ACTDIK008) Recognising that numbers, text, images, sounds, animations
and videos are all forms of data when stored or viewed using a digital
system
• (ACTDIP009) Using software to sort and calculate data when solving
problems
• (ACTDIP010) Explaining what the problem is and some features of the
problem, such as what need is associated with the problem, who has the
problem and why
• (ACTDIP010) Designing and implementing a simple interactive digital
solution using a visual programming language

UK NATIONAL CURRICULUM: COMPUTING (KEY STAGE 2)


• Design, write and debug programs that accomplish specific goals, including
controlling or simulating physical systems; solve problems by decomposing
them into smaller parts
• Use sequence, selection, and repetition in programs; work with variables
and various forms of input and output

7
NAME: ______________________________________________________ DATE:_______________________________________

FORMATIVE ASSESSMENT: SUSTAINABLE FARMING

What is a dataset?

Why is land optimization an


important thing when farming?

How can AI help developing


countries increase their food
production?

You might also like