Open-Ended Project

At this point in the class, you have gone in depth with a multi-week project in both mobile robotics and computer vision. Now that you have built these core skills, you are capable of pursuing a large range of projects in the computational robotics space. The purpose of this final project is to give you an opportunity to exercise, reinforce, and expand your skills in this area.

Teaming:

I am recommending that you work in teams of two or three, but I am willing to allow larger teams under the following conditions:

1. There is a sub-team structure that allows various aspects of the project to move forward independently. Integration will happen periodically throughout the rest of the semester.
2. For your final deliverable, you will need to have a presentation that is twice as long.
3. You will need to turn in two separate project writeups (one for each sub-team).
4. You will need to do something really ambitious and cool!

Differences Between this Project and Previous Projects

One obvious difference between this project and previous projects is duration. This project will be five weeks long (not including the Thanksgiving week, which would be a sixth week). The additional duration provides some opportunities and challenges.

Opportunities:

• You may want to consider utilizing additional hardware above and beyond the standard Neato platform. This could come in the form of a different robot, a new sensor, or adding a new actuator (e.g. a robotic arm).
• You may want to consider adding real-world context to your project. For instance, perhaps you could create educational or outreach activities for elementary school kids that involve learning about robotics. Another idea would be to get a robot to solve a tough real world challenge (like searching for survivors in a disaster scenario).
• Your project might integrate a significant work in both mobile robotics and computer vision. With the added time, you have the opportunity to really integrate across the two major content areas of this class to do something awesome.
• Your project might integrate with other projects that you are pursuing at Olin. The additional time makes a meaningful integration possible.
• You can take a deep dive into a particular algorithm. With more time, you can explore various algorithms for solving a task in a much more systematic and comprehensive way. You might also be able to create your own implementations in ROS or OpenCV of techniques that currently do not have widely available open source implementations.

Challenges:

• With more ambitious projects comes more risk. You should have contingency plans in place if your initial plan proves infeasible.
• Longer projects are tougher to undertake from a scheduling / project management standpoint. You will have to be diligent about continually making progress on the project despite the long duration of the assignment.

Design Review:

On Tuesday December 2nd and Thursday December 4th we will be having final project design reviews. These reviews will take place at the beginning of class. The project teams will be divided into two separate groups. Teams working on similar projects will be placed in the same group (the groups for Tuesday have already been selected, and on Monday we will select groups for Thursday).

Each project team will take turns presenting to the other teams. Each team will have up to 30 minutes to discuss their project with the other teams. However, it is okay if your review runs short of this time. The purpose of the design review is to gather useful feedback and suggestions from your peers. As such, the review should not be structured as a formal presentation of what you have done so far (which would be more similar to a UOCD design review), but instead should be designed to engage the audience to resolve issues and move your project forward (more like a SCOPE design review). It is completely up to you how you will scaffold this engagement. Here are a couple of possible structures for the design review (this list is by no means exhaustive and several of these could also be combined):

1. Present what you have done so far via slides, and then end your presentation with a list of questions / issues that you want the audience to engage with.
2. Engage the audience in an ideation / brainstorming activity to move your project forward.
3. Show your code base for the project and ask for suggestions on how to better structure the code.
4. Ask for feedback from the audience on your process of working together as a team and how it could be improved (e.g. you might ask for advice on integration, unit testing, task tracking, etc.).

It is up to you to determine how to use this time productively. Your peers are likely facing similar issues as you are (both from a process and a technical standpoint), make sure to take advantage of their expertise. To do so, you should consider who your audience is and what relevant background information you will need to present to give your audience the tools to help you move your project forward.

Robot Expo

On December 18th we will be having a Robot Expo. I am planning to invite the Olin community to come to this event. The structure of the expo will be 50% presentations (probably about 15 minutes per team depending on how many teams we end up with in each section), and then 50% demo / poster session. The demo does not have to be a live demo (although a live demo of some aspect of the system would be nice). You should be prepared to stand by your poster and talk about it with other students and Olin community members as they come by. So that you get a chance to see other people's posters only half of your team needs to be by the poster at any point time (allowing the other member(s) to go view other teams posters / demos).

Intermediate Deliverables

Project Proposal: During the class after the kickoff session, I will be meeting with each team to discuss their project proposals. Before class, you should e-mail me a copy of your proposal with the following content:

• What is the main idea of your project? What topics will you explore and what will you generate? What is your MVP? What is a stretch goal?
• What are your learning goals for this project?
• Given that the project will be four weeks long (five if you count thanksgiving), describe what you'd like to have complete at the end of each week.
• What frameworks / algorithms are you planning to use? (if you don't know enough yet, please outline how you will decide this question during the beginning phase of the project) Another way to think of this would be, what content that we have learned in CompRobo will you likely be bringing together in this final project?
• What do you view as the biggest risks to you being successful on this project?
• What might you need from me for you to be successful on this project?

Final Deliverables

• Your code
• A writeup (see below for format)
• A poster for the robot expo
• A demo (either a live demo or video) that you can show to people as they tour around the robot expo.

Writeup: in your ROS package create a file to hold your project writeup. Any format is fine (markdown, word, pdf, etc.). Your writeup should touch on the following topics:

• What was the goal of your project?
  • How did you solve the problem? (Note: this doesn't have to be super-detailed, you should try to explain what you did at a high-level so that others in the class could reasonably understand what you did).
  • Describe a design decision you had to make when working on your project and what you ultimately did (and why)? These design decisions could be particular choices for how you implemented some part of an algorithm or perhaps a decision regarding which of two external packages to use in your project.
  • How did you structure your code?
  • What if any challenges did you face along the way?
  • What would you do to improve your project if you had more time?
• Did you learn any interesting lessons for future robotic programming projects? These could relate to working on robotics projects in teams, working on more open-ended (and longer term) problems, or any other relevant topic.

A Potential Class-wide Competition

A problem of huge importance in robotics is searching for objects of interest in a novel environment (i.e. one the robot has no experience operating in). One particular area of application of this technology is creating robotic first responders that can identify survivors in the case of some disaster (e.g. an earthquake, bombing, hurricane, etc.). I know several teams are interested in doing some sort of class-wide competition. What I think would be a fantastic culminating experience for the course would be to have teams work on the problem of having the Neatos search for objects of interest in the old star center.

The basic idea would that the robot would be placed at the door of the old star center, and the robot would have a fixed amount of time to search for several objects whose appearance would be known to the teams a priori (these objects would probably be colored in distinctive ways to aid their identification). At the end of the time, the robot would have to produce an estimate of where each of the objects was in the environment. The team's final score would be determined using some yet to be determined scoring metric based on the accuracy of the predicted object locations.

I like this task for several reasons:

• There is a compelling real world application
• In order for teams to be successful they would need to make use of many of the techniques learned in this class (both in the mobile robotics and computer vision spaces).
• It would be fun, and if several teams work on it there would be a lot of opportunity to share ideas / insights.

Anyway, I definitely don't want to mandate this project, but if you are already thinking a competition-like project would be something you would be interested in, then I think this is a great candidate to consider.