Learn AI with a robot

Amitabha
Amitabha
Dec 15, 2019 · 7 min read
Price shown above is a sample illustrative price. Order at https://robotics.thinkific.com/courses/aiwitharobot

Why?

For most of my career, I have been an engineer solely focused on the performance aspects of computer operating systems. Three years ago, I started driving an analytics and AI intensive feature for an emerging product at my current employer. Learning the subtleties of Artificial Intelligence (AI) was an uphill struggle for me. Luckily, at the same time, I was also a beta tester for a robotics company called Anki, who were releasing their first full fledged toy robot, the Cozmo. Along with Cozmo, Anki also released a full fledged SDK to provide the ability to program Cozmo using python. Programming the Cozmo was a great enabler for me to learn AI, as gradually I could learn and comprehend how AI was widely used to make Cozmo both technically and emotionally smart.

My experiences of learning with Cozmo motivated me to design a online course on learning AI with the help of a robot. Three years down the road, Anki released another robot called the Vector. Vector is an even more advanced robot than the Cozmo, because it is fully autonomous, can locate a charger and charge itself, and can interact with a computer program via WiFi without any additional hooks (The Cozmo SDK required one’s phone running the Cozmo App to be connected to the computer running a program in order to send instructions to the Cozmo robot. The Vector SDK allows the computer running a program to directly send instructions to the Vector robot via WiFi.). Therefore, while Cozmo’s brains lies in the phone, Vector has its own brain and is a complete robot. I found that Vector and the Vector SDK was an even more powerful companion for learning AI than Cozmo.

Thus began the adventure of designing an online course on how to learn AI with the Anki Vector robot. “Learn AI with a robot” is designed with a goal to teach you five major concepts in AI with the help of Vector. These five big ideas are being developed by AI4k12.org (An organization dedicated to building an AI curriculum for schools in the United States). A poster describing these five concepts is provided below. “Learn AI with a robot” tries to explain each of the big ideas defined by with the help of illustrations and demonstrations using the Vector robot.

Who is this course for?

The prime target audience of the course are upper middle and high school students who want to get a flavor of AI and the different aspects/ dimensions of AI. The course is also a great starter for those who want to get into the hobby of robotics, or those who are fascinated with how robots are engineered with AI. The course will also be invaluable for non computer science/engineering major undergraduate students who might be curious about how to use AI in their domain.

The course assumes basic middle school science concepts. While you would benefit if you own a Vector robot, we do realize that buying a robot might be prohibitively expensive for many school students. (Unfortunately, Anki went bankrupt in May 2019, but its robots can still be purchased in the market and on Amazon at a cheap price.) So the course is designed assuming that you don’t own Vector. With every chapter, there is a ton of accompanying reading material to help you deep dive into the subject. If you want to get in tune with the latest research, the links will provide you with the right pointers.

5 key AI ideas that should be taught in K-12
5 key AI ideas that should be taught in K-12

How is this course different from other AI course, many of which are free?

There are abundant free resources on Coursera to learn AI from. My favorite teacher for AI is Andrew Ng. There are many courses by him on Coursera. I have been a student in many of these courses, and he is a fantastic teacher. However, many of these courses are tailored towards undergraduate computer science students, and expect high prerequisites. The courses also focus more on theory. There is also an exhaustive list of curated resources maintained at AI4K12.org, many of which are particularly oriented towards K-12 students.

We distinguish ourselves in the following two ways:

  1. Video Demonstrations: Learn AI with a robot aims to show you more practical aspects by including lots of video demos based on the Vector robot. I think the video demonstrations are really helpful for students to visualize and comprehend the theory of AI. So the first value I expect you to get is to see the video demonstrations along with open-source code that is behind these demos.
  2. Comprehensive Organized Resources: If you are too busy to hunt for resources for yourself, you would definitely find our list of reading material very valuable. I post every useful link that I find in the reading material appendix of each chapter, along with why the reference is valuable. The list of references will grow over time. If you buy the course, you have lifetime access to this reading material.
  3. Personalized content: This course is your chance to learn from me. So, if you like my posts on Programming reports on Medium, or enjoy my Youtube videos, this course is really good for you.

Below is an outline of the course curriculum and what you can expect to learn from this course.

Course Outline

  • Introduction (Available for free preview). Provides an outline of the corse and what we expect the students to learn. 6 minutes of video.
  • 3 interesting applications of robots (Available for free preview). Discusses the use of robots in warehouses, exploration, and in automation. 17 minutes of video.
  • Perception. First two units are available for free preview. Discusses how Vector uses sensors and algorithms to perceive the work around it. 28 minutes of video.
  • Representation and Reasoning. Discusses the various data structure that Vector uses to represent its knowledge. Considers how robots compute paths. 20 minutes of video.
  • Learning Part 1 — Linear Regression. Introduces the concept of how machines can learn by discussing linear regression. 23 minutes of video.
  • Learning Part 2 — Convolutional Neural Networks. Introduces the most hot topic in AI: Deep Learning by demonstrating how Vector can be taught to learn the human sign language. 54 minutes of video.
  • Human Natural Interaction. Introduces the hardest challenge in robotics… making robots behave like humans. Discusses animations that are programmed in Vector in great detail. 19 minutes of video.
  • Societal Impact. Discusses the impact that AI and robots will have on society. Attempts to balance the discussion between good uses and risks of technology. 18 minutes of video.
  • Projects: This part of the course is optional. We have a couple of fun and interesting projects you might want to attempt, especially if you are in possession of Anki Vector.

That is 3 hours of packet content which will give you a great introduction to AI with practical use cases. If you already are familiar with AI, this course will make you an expert on how AI is used in robotics. So hop on the bus, and let’s have a great ride. In case, I still didn’t convince you, the following video from the first chapter of the course just might.

To access the course, please visit https://robotics.thinkific.com/courses/aiwitharobot

How can I buy this course?

First of all, we would like you to remember that the first two chapters, and two units of the third chapter are available for free. That comes to approximately 25% of the course.

I would be honored if you choose to continue and learn the remaining material with me. Here are some ways to buy the material.

  1. At https://robotics.thinkific.com. This is the preferred way and gives you instantaneous access. You can pay using Paypal which supports all major international credit cards.

However, I do realize that many international students lack the ability to pay is US Dollars for many reasons. So here are other payment options. Remember that none of these are instantenous, that is they don’t give you immediate access, because I need to verify the payment. When you pay with any of the other means, you need to include your email address in notes of the payment. You will receive an email from me in 1–2 days. Once you send a reply conforming your identity, your email address with be added to our system, and you will receive an introductory mail on how to access the course.

2. Venmo: Please send me $19.99 (or the current listed price at https://robotics.thinkific.com in case a discount is being carried. Please note the same that you saw a discounted price and I will honor the transaction) using Venmo.

3. Indian Rupees: For my students in India, you can may in Indian Rupees. You must be a resident in India to avail of this price. The price in Indian Rupees is 999.00 Please use the following link:

This link will allow you to pay with any major Indian payment systems such as Netbanking, RuPay cards, etc.

Can you tell me about yourself?

PS: A bit more details about myself. I have 15+ years in the software industry with stints at VMware, Oracle, Sun Microsystems, and Tavant Technologies. In-between I did a PhD in Computer Science from University of California, Davis. For the last 5 years, I have been applying different AI technologies for multiple use cases for my current employer. I have published over 30 peer reviewed scholarly articles, and have been awarded 5 patents. If you want to know about me, please visit my LinkedIn profile.

Programming Robots

Learning AI by programming robots. Continuing at https://programmablerobots.substack.com