Friday, June 24, 2016

Grand Challenge Design: Simulation Concept

This fall, I will teach a new year-long 90 minute/day (2 credit) course, Grand Challenge Design.  The big picture goal is to explore major world problems (inspired by the National Academy of Engineering's Grand Challenges and the Future Problem Solving Program) and create working solutions using the technologies that are fueling the "Internet of Things", physical devices that become "smart" by connecting to the internet.

The initial plan was to spend the fall working on tutorial-style projects to ramp up student skills with web application design, electronic circuits, and using the Raspberry Pi computer to connect them.  Then the spring would be used to connect student teams with local businesses, non-profits, and individuals who could work with students to identify problems and test out their solutions.  Based on this plan, I was able to recruit a class of students to sign up and excite a number of people around the district and community.

As I started working on it and talking through details with my wife, we both thought that the early part would get a bit boring.  The projects sounded neat, but that long ramp-up period in the fall was super technology-focused and didn't integrate the real heart of the course -- the Grand Challenges.  This also meant that the course would be doomed to forever being a just another tech course for techy kids.  I think that ALL students need to study global issues and that ALL students need basic proficiency with the technologies that will power the future, so this wasn't going to work.

The new plan is to immediately drop students into an immersive role-playing simulation on the first day of class.  The concept is based loosely on John Hunter's "World Peace Game".  At first glance, it looks like the awesome board game "Risk" on steroids.  Nations have cash, natural resources, military personnel, and various capabilities.  He introduces 50 or so conflicts that the student-run nations need to solve.  In order to win the game, all conflicts need to be resolved and all nations need to end up better off than when the game started.  (See these example crises from one teacher's version of the game.)  However, student nations can act in their own self interest and have no obligation to listen to their peers.  Students get very immersed in the game and experience deep learning.  The best explanation I found online was the video below (focus on minutes 10-18).  He also has a TED talk that goes over the high level purpose of the game with some duplicated footage.  I just bought his audiobook and look forward to listening for advice on running meaningful role play / simulation activities in the classroom.

The goal of Hunter's game is to help students experience the challenge in creating world peace.  The simulation takes anywhere from a week to a couple months depending on how much time is spent per day.  In Grand Challenge Design, the goals are a bit different.  While diplomacy is an important part of any cooperative/competitive role play, my primary goals are that students (1) learn more about major world problems and (2) learn how to use the IoT technolgoies we need in the second half of the year.  In addition, this scenario needs to last 90 minutes/day for 3-4 months.  To that end, I think that the challenge scenarios need to be a bit more complicated and require student research before reasonable solutions are created.  Students will need some kind of incentive to bring in outside council from an expert in the topic area to discuss the issue with them (in person or digitally).  The game will also need to be challenging enough that it cannot be won with diplomacy alone -- technology will need to counter what would otherwise be an impossible scenario, just as it does in real life.

The major challenge areas I want to focus on are healthcare, the environment (primarily water quality/supply), infrastructure, the food supply, and security.  Naturally, these topics are heavily interconnected: farming has a huge impact on fresh water, much of our safety is based on the quality of infrastructure, health is dependent on the environment, disease tracking and security monitoring are both positive but at odds with personal liberties, etc.  The goal is to create a simple enough game that introduces problems in these areas to the student teams, a set of city-states sharing a fictitious island.

As part of the solution-finding, teams will need to develop smart devices such as water pollution monitors, intelligent soil monitors, stress monitors for bridges, biometric monitors, and cameras with facial recognition.  These smart devices will only be useful when they feed their data to a server where it can be aggregated with data from other devices.  From there, students could create simple dashboards to monitor data manually, pattern monitors that detect major changes and alert them to changes, and systems that directly take action without human intervention (such as telling a tractor where to fertilize or opening a dam to maintain water levels).

Obviously, this is quite the design problem to make this all work.  Due to the tech skill-building required (most students are starting from near zero with our tech, even those who are fairly "techy" kids), I want the game to start off with a gradual introduction to server APIs and Raspberry Pis.  Maybe the monetary system could be all electronic and if teams wanted to spend or exchange funds, they would need to create HTTP requests to an API that I write with a few parameters to use money.  I want to make sure that everything that students need to know gets tied into the game narrative.

I also want to make sure that every student in the class learns all of the technologies, not just divide and conquer (meaning that the skill gap between those with more or less tech experience widens).  I might require that all citizens of a nation check their personal health stats from a server I setup or build a health monitoring device that alerts them to changes in vitals (based on my server).  Other challenges, such as environmental monitoring, could let teams specialize into those who setup sensors and electronics and those who collect and process the data.  Though students are put into teams by nation, many projects will be inherently international and lead to new teams of students collaborating.

As one example to fully integrate the smart devices into the simulation, I could have teams put bowls of actual water in their city with real sensors in them.  That way, when I "pollute" one of the water ways, a real sensor will have to detect it, and hopefully it will cutoff water to their citizens before mass sickness takes place.  In another example, a few people will get sick at a time with a clear disease vector developing.  If the nations pool their health data and have someone monitoring it as the CDC representative, a major outbreak can be averted.  On the diplomacy side, if the richer nations don't help the poorer nations track this kind of health data through their own funding, citizens of all nations could suffer as outbreaks get started in untracked nations.  In another example related to security, nations could setup cameras to recognize faces and alert authorities if people on the watch list are found, but then new conflicts around public opinion on privacy would emerge.  In this way, the game prompts student research into problems, which leads to action and new device building, and the solutions prompt further problems and discussion.

In order to make sure that global issues research is a major focus of the course, I don't think problems should all be released simultaneously.  As an example, I could start by focusing on healthcare for a few weeks and gradually move towards the environment and so forth.  This way, students can take time to research one topic well and develop ideas and solutions to the crises.  I can also look for experts in the community to come in and join us for class during various periods of the semester.  The downside to this approach is that things get a bit too teacher-controlled, forcing me to more carefully manage the pace and flow of the game.  This control could help me make sure that all teams always have at least one critical challenge in front of them at any given time.  I could also assign temporary new international roles or move students around as needed.  In Hunter's game, he generally gives out all problems at the start and keeps teams static.

Besides all of the ways that the game can go wrong or fall out of balance, one challenge I worry about is too much success: students wanting to continue the game rather than work with real customers once they are well trained on the tech and exposed to a variety of challenges.  Real customers means that you have to talk to new people, go off site, and deal with the complexities of real users in their real environments.  External people and environments don't always play nice with electronics.  I don't want the most awesome part of the course to be a let down, so it will be important for potential clients to come into class to take part in the game and see what students are capable of thinking about and building.  That way, they should have a better sense of how to engage students in their own contexts and get them excited about their challenges.

As this post makes evident, this idea is still half-baked at best.  I need a ton of feedback to really develop the simulation into something that makes sense, has a clear objective, and incentivizes the right behaviors all while still being fun.  Given that 100% of this class will be open and freely distributed, the game will need to make sense to others, not just me, so they too can use it.  Call/text/email/FB message/Tweet me up, or leave lots of comments below.  Thanks!