I just drove most of the way across the country (from upstate New York to New Mexico, just shy of 2500 miles). This gave me ample opportunity to ponder my GPS, what its underlying algorithm is, how it comes up with routes, and what the interface is that I’d like it to have. It occurred to me that a GPS could be used as the foundation for an interesting classroom exercise. And, since lots of people bemoan the fact that nobody can read maps anymore, this exercise can promote good old-fashioned map reading along with thinking about computers and computing.
Here’s a basic idea, though it certainly could be modified depending on the age of the students and how vast an area you wanted them to consider.
1. Give the students a map of the relevant geographic area. For younger students this might be just the neighborhood around the school. For older students you could work at the city, state, or national level.
2. Ask the students to use the map in order to generate multiple routes between two points. Based just on what they see on the map (using the legend to determine road type, etc.), have them predict which should be the shortest and which should be the fastest routes.
3. Then fire up a GPS and see what it says. Maybe do some advance work by actually driving the GPS’s chosen route, see how long it takes. Try some variations based on what you know as a driver (or a walker, for those in urban areas).
4. Work with the students to try to determine how the GPS is making its decision about the “best” route, what the algorithm might be, what factors it is capable of considering, which factors it ignores. For example, the longer route with fewer traffic lights will be faster but the GPS (mine, at least) always chooses the shorter route with traffic lights which invariably takes more time.
At the upper level, one could have students plot out a few cross country routes, compare the Google map options to the GPS route, research information on rush hour traffic in major cities, and consider the option of smaller roads that avoid urban areas. It’s a relatively simple scenario which turns out to be rich in problem solving and algorithmic opportunities.
When I ignore my GPS’s directions, I’m particularly intrigued with how long it takes for me to get far enough along my chosen route that the GPS will finally stop trying to get me back to its chosen route. What I really want is for the GPS to initially present me with options (the way Google maps does) so that I can tell it whether I want the longer scenic route, the urban centers route, etc. But for now I’ll just relax until I have to get back in the car and make the return trip (taking a totally different route home!).
Valerie Barr,
CSTA Computational Thinking Task Force Chair