"Setup"

Pella Externship Days 16-17 The last 2 days have been aimed at a 2nd version of the Optimization Program.  The first few attempts at running the simulation showed me that there was some debugging to do.  I've incorporated the new constraints and variables, but the big breakthrough came with a different approach to the regression equations.  Instead of using many equations to relate one variable at a time to the outcome (which produced some extremely erratic solutions), I tried out the Data Analysis program in excel.  The program allows for multiple variables to be combined to predict a specific outcome.  I'm still a little concerned about the correlation, but I got the relationship from a 10% or less accuracy rate to 80% or better predictability. My "Ah-ha" teacher moment came today as I started on a User Guide for my program.  In a conversation with Mike yesterday, he asked that I put together some instructions to accompany the program.  That way anyone trying to use the program would have a general understanding of how it was structured and how to utilize it.  I spent some time today trying to remember and document what I was thinking as I designed the opt. tool.  It was difficult at first, but as I got rolling with it I realized how valuable it was, not just for the user, but for myself and my understanding of the relationships involved. I use written reflection in the classroom often, but it is good to be on the other side of it once in a while.

"2nd Wind"

Pella Externship Days 12-15 The past few days have been great.  The initial excitement of the project is back.  I spent about 3 hours last Thursday going over what we had discovered by working the data so far.  It was great to get some feedback on the direction I was heading and some suggestions on how we might want to incorporate the models I had come up with.  We also discussed some new components of the Optimization process that we should consider including.  New twists, like being able to purchase some parts instead of producing them from the lumber plant, have given me a different perspective of how the process could be better managed.  Other components, like the demand for certain parts and how much raw lumber inventory is available, are going to play a key role in how effective the program can be on a daily, weekly, and monthly basis.  In addition, we discussed more practical views of overproduction in the plant. All of our discussions have provided some clarity as to the final product they hope to have at the end of my time here.

"Try, Try Again"

Pella Externship Day 7-11 I spent the end of last week trying to pull some useful equations from the past several years of data.  I'm starting to understand the difficulty involved.  Any best fit line that is calculated is hard to justify using for optimization.  When you look at the predicted value using the equation, there is often a huge variance from the actual results produced.  Yesterday and today I looked into another option for predicting the amount of lumber recovered.  The lumber plant has some test results from specific mills that they have collected and studied over the past 2 years.  While there are a relatively small number of samples taken for each mill, they were isolated and specific.  In addition, a portion of each of the past 4 days has been invested in designing different versions of the optimization tool.  With some luck, I think we will be able to combine our conclusions from the data into a functional program. In terms of lessons to bring back to the classroom from this week, I have caught myself several times searching the Internet for answers to my programming problems.  As a math teacher especially, it is easy to dismiss the importance of inquiry in the classroom.  I need to find opportunities for my students to find solutions to challenging problems independently.   With the pressure to cover everything in the curriculum, it is easy to forget the importance of students creating their own connections and being given time to struggle with difficult problems.  I need to remind myself that the time invested early on in the school year to build confidence and independence will allow for a much smoother 2nd semester.

Pella Externship 2011- Week #2

Pella Externship Days 3-6     It has been a fun past few days as I jump into some of the data.  I've really enjoyed getting the figures compiled and sorted so they are easier to work with.  I hope we are able to find some regression equations of value.  It almost seems like the harder one looks at the number or tries to manipulate them, the less clear the outcomes become.  My concern for the project right now is that our equations may lack the correlation to make them useful as prediction models.   In addition to picking up some programming tricks in excel, I've had to learn a few things about cubicle work as well.  As a teacher, I take for granted how nice it is to stay moving all day.  On more than one occasion my body has had to remind me to blink as I focus on the computer screen for hours at a time.  I never thought I would come to appreciate the chaos of the hallway during passing time.   The lesson of the week that I will take back to the classroom is the importance of active learning for the students.  Sometimes I forget that I have the luxury of walking around the classroom and interacting with students on an individual or small group level.  For the most part, students are sitting in a desk all day.  While I try to find a way for students to interact with one another often, I think I can bring in more opportunities for kids to get out of their seats and collaborate.

Pella Externship 2011

Summer Externship - Day 1

Pella Corp.

 

The first day of orientation at Pella Corp. has gotten me very excited about the summer project.  We began with a tour of the Lumber Processing plant.  As we follow the lumber from its entry in to the warehouse through the entire processing center and on through final assembly, it becomes clear that there are a lot of variables that will be hard to isolate in the optimization process.

 

As I talk to Jeff and Mike, my primary contacts, I start to get a feel for some of the methods they have tried to collect usable data.  I start to realize that in my math classroom I present students with "Real World" challenges that provide easily attainable data and nice and neat solutions.  In the actual Real World, the data isn't always so black and white.

 

As we finish the first day, we talk about some expectations and possible directions we can head with the summer project.  Ideally, we would be able to create an optimization tool in Excel to help make decisions about what lumber is purchased and fed into the plant to raise yield and lower costs. 

 

 

 

Summer Externship - Day 2

Pella Corp.

 

The morning of my second day is spent looking at data that has been collected over the past few years.  Jeff and Mike do a great job of explaining how the data has been summarized.  Overshadowing the data is the fact that multiple sources of lumber are mixed daily, making it difficult to assign yield values to particular mills.  Just before noon, we meet with some of the team that has been working on the statistical data.  It is agreed that we need to dig in and try to identify usable conclusions before an optimization tool is realistic.  The remainder of the afternoon is spent getting comfortable with the data .

 

I can tell that my experience will be challenging and rewarding this summer.  I've already seen a need to bring more ambiguous problems into my classroom.  Can't wait to see where this will lead me.