What schools could learn from cab companies

I spent the first year of my professional career as a teacher.  At the risk of sounding arrogant, I thought I was pretty good--I had good rapport with students, I tried to align my teaching with learning objectives, and I believed that I applied good pedagogical practices in my classroom.  It's probably pretty easy to spot the problem in this assessment--they are all my own personal perceptions of my ability and performance, and very subject to bias and inaccuracy.  This has bothered me lately and I've wondered how good a teacher I really was during that year.  As I look back on those experiences, two really important things seem to have been missing:  feedback and focused effort to learn from mistakes.  

While I tried to regularly evaluate my own teaching, like most of us I probably overestimated my abilities and was likely unaware of many of the mistakes I was making.  On a few occasions (three, that I can remember) I was observed by another teacher and then given some basic feedback at the end of the class session.  It's fair to assume that both my own evaluations and those of others might have had some slight improvement on my teaching.  But, in retrospect, I don't think I was much better in June when the school year ended than I was on the first day of school in September.  In fact, I probably developed some bad habits, got too comfortable with the role, and stopped doing some of the little things that make a big difference in one's teaching ability.  In short, I may have even been worse.  

At the same time I was having these depressing thoughts, I was reading Traffic, by Tom Vanderbilt.  The book uses driving patterns and habits as a context for exploring human behavior.  He tells a fascinating story about how cab companies and limo services have helped improve their drivers' performance using a technology called DriveCam.  DriveCam installs cameras on the rearview mirrors of cars that continusously buffer images (like TiVo) of what is happening both inside and outside the car.  Sensors monitor various measurable forces and when a "trigger" is detected (sharp turn of the steering wheel, significant decrease/increase in speed, etc.), the camera records ten seconds of footage both before and after the trigger.  This footage is then sent to a database and may be reviewed with the driver in an attempt to correct mistakes and improve safety.  Although it probably ruffles some feathers of drivers who feel like their privacy is being invaded, the bottom line is that transportation companies have a lot to lose when their drivers don't drive well.  What's more, this is great learning.  Drivers can view actual footage of their driving, spot mistakes and fix them, and focus on small elements of the driving performance that seem to make a big difference in achieving good outcomes, namely safety (this isn't unlinke the way elite athletes use taped performances to improve).

This left me wondering why something similar couldn't happen in classrooms, particularly those classrooms led by novice teachers.  We spend millions of dollars to equip classrooms with new technologies that are touted to improve student learning and increase engagement.  And, while ipods and laptops can help, the core factor influencing student learning still seems to be the teacher.  It seems fair to ask why technology can't be used to imrove core teacher practices that would have far-reaching impact upon student learning.  There is some work being done in this area (see the work of Peter Rich and this pilot project at the University of Central Florida), but it  seems to be on the periphery.  

What would happen if there were cameras installed in classrooms that could capture real footage of teacher performance.  Are there "triggers" that we would want to focus on?  What would they be?  Something like this would seem to fill a gap in current teacher development practices.  Teachers would have the opportunity to really see themselves teaching (as opposed to their perception of their teaching or someone else's interpretation), use mistakes to improve, and focus on critical parts of their performance that have been shown to lead to significant improvements in student learning.