Tuesday, July 14, 2009

ChMS - Lessons from week 1 of check-in

Sunday night we launched check-in for our 7 community. We launched it at 7 for several reasons:

(1) There are only about 600-1000 adults that attend 7, and most of those are younger resulting in fewer kids going into our program.

(2) The fact that most are younger means they are less likely to be antagonistic towards technology.

(3) It is isolated from the rest of the communities, being the only one meeting on Sunday night. As such, it is a perfect place to deploy first.

Here's what we did wrong:

(1) We ran out of time. This is probably true of almost any deployment, but we got bit by it. Not hugely, but we did get bit. We were so focused on the mobile kiosks that when we went to setup the desktop kiosks for nursery and pre-school, we realized we didn't have enough wiring in place to support two kiosks, so we had to settle for one for the first weekend. Unfortunately, the fact that 7 is a younger crowd meant that the load on these areas was pretty high, and having only one kiosk was a huge problem. When someone wasn't in our system properly, a big line would build while we were entering them in. Although we had never planned for there to be only one, it definitely reinforced this principle: Always have at least two kiosks in every location where people are checking in.

(2) We tested the family registration app, but not with lots of different addresses (or invalid ones). This doesn't seem like a big deal, but by request we had included geocoding on the spot for new addresses. Unfortunately, when an address wouldn't geocode, the record wouldn't save, creating a bit of a panic when you have a huge line of people (see point 1) and the record won't save allowing you to check-in the family. This was a small mistake, but it bit us big time. The reason the record wouldn't save is that the geocoding was timing out on the request, which took a long time. I honestly don't know how we could have known to do this differently.

(3) We didn't educate the kids team enough on how check-in should work. For example, they gave us very precise ranges for kids rooms (0-6 months, 7-12 months, 13-18 months etc.) which seems like a great idea, but doesn't allow the system to "load balance" rooms resulting in a few very full classrooms offset by some rather empty ones. Always provide age overlap so that the system can balance rooms for you.

(4) We didn't hound the parents for good data ahead of launch. We had LOTS of kids with no birth dates in our system. This resulted in a lot of time wasted while we added birth dates with kids in line. Take time on the front end to fill in holes in your data.

The elementary team was extremely happy with the new check-in system. Even though we were dealing with a lot of missing information, check-in went very fast because this system is so much faster than our old processes.

The pre-K team wasn't as happy because of the issues we listed above. But we are fixing those for week two. We will go live in all of our communities once we feel that 7 is working as it should. On Monday we had a debrief meeting in which we laid out a schedule to go live in August for everyone. Prior to that, we are going to ask parents to fill out cards that have information like all of their kids names and birthdays etc. If even half of them do that for us, it should help us immensely.

On the whole we are very pleased with the system so far. Many, many kudos to Nick and Jason at CCCEV for their work on a check-in system that allowed us to go live with an all mac check-in system. To our knowledge we may be the only church in America doing automated check-in without windows client software. Arena let us do that, and that is very cool.

Joel

Sunday, July 12, 2009

ChMS - Check-In Progress

I'm sorry about the gap in blog posts. There has been a tremendous amount of stuff going on, and I've been too busy to read my email, let alone blog...

We made the decision a couple of months ago to pull the trigger on children's check-in. This is something we have wanted for at least nine years, so it is quite exciting to finally be making it happen.

The first step was software, and we found a solution that would work for us in the check-in module written by CCCEV. I blogged about this here. Next up, we had to create a replacement for the family registration windows app since we were planning to do this as an all mac solution. Daniel wrote a replacement for this as an Arena module, and we involved the Arena community in its development to insure that we benefited more than just HDC with it.

The biggest thing for us is being able to do this as an "all mac" check in solution, allowing us to work with equipment and software that we know well. Once we made the decision to move forward we had to write the family registration module, find a kiosk and get it ordered, and get the equipment in place. That's when we hit our first snag: We saw lots and lots of kiosks online, but we could not find what we needed, so we ended up building our own. The kiosks we looked at failed to meet our needs in several ways: (1) Most were not mobile. We had to have them on wheels. (2) Most did not have space for the printer. We wanted to print the labels at the kiosks. (3) None of them came equipped with macs, although many came pre-loaded with pc hardware we could not use. (4) Most of them were much larger than we wanted. We have to store these things, they need to be small. (5) None of them had a provision for a keyboard to be used only some of the time. It was either no keyboard, or a keyboard out all the time.



This is the final product of our design. The monitor is an ELO Accutouch 15" LCD. We tested two other types of touchscreen (Intellitouch and Caroltouch) from ELO and this was the favorite of our staff. Below the monitor you can see the tray that holds the keyboard. It recesses into the kiosk. The idea here is that people come up and primarily interact with the screen as a touch screen. The keyboard is utilized by volunteers when these need to be used to register new families or make changes to the data in our database. For the most part the keyboards remain hidden. Then the printer, a Zebra GK420d.



The back of the kiosks show the way we have assembled everything. There is a mac mini mounted on the side, a linksys 5 port ethernet switch (powered by Power over Ethernet) and then a power strip. What this photo doesn't show (it wasn't installed yet) is a panel on the back along the bottom that provides external plugs for power and ethernet. So the volunteers only have to take the 12' cables and plug them in on the outside. There is no stretching of the power strip cable, no stress on the switch inside etc. Everything is self contained. We also found USB powered fans to keep everything cool. The best part about them is they only create 19db of noise: they are virtually silent.



Then we had to start setting everything up. This means a serious assembly line of stuff. Here you see 16 mac minis being installed and tested. Below, power strips, mini shelves and ethernet switches await installation.



The kiosks are not quite finished. They are due to be wrapped with a vinyl wrap that includes kids graphics and such to make them blend into our environment better, but we stained them ebony to minimize the showing of dirt on exposed areas, as well as to make the kiosk disappear behind the printer.



This is the first batch of kiosks, eight of them, ready to go for this weekend. We are implementing check-in at our 7 community, which has a much smaller group of kids than our Harbor community. We figured that 7 is the perfect place to test and work out the bugs. We go live tonight, I'll let you know how it goes tomorrow!

Joel