Astute followers of my blog know that I am fashionably late to this upload, which is especially eggregious considering I told you to “mark your calendars” for July 25th last time. I’d like to take this opportunity to apologize to all my adoring fans. You all really deserve better. I am here now, though, and ready to deliver premium content!
Technical Updates
As you may have guessed from the title of the blog, the road to success on our main project has been a bit of a bumpy one! One of the saddest bumps was when I broke what we later realized was our only OLED display with the specifications we needed for our incubator readout. This forced us to switch out to an LCD display we had available, scramble for a new display/button housing, improvise different wiring, and revise code.
I should not have pushed with all my might to attach this display to the 3D printed housing. The more you know.
Although it was a frustrating setback at first, my destruction of the old screen actually led to several improvements in our device! The larger replacement screen makes the user interface more intuitive, and another look at the code led to several ease-of-use improvements that made for a more robust device, like an initial welcome screen and improved button delay for better responsiveness. Another improvement that comes with the replacement screen is a reduced computational load on our microcontroller. The new screen has its own processor specifically for controlling the display, freeing up our Arduino for intensive tasks like calculating the baby’s pulse rate or potentially blood oxygen levels.
The new LCD replacement screen saved our lives! It’s also harder for me to break (so far, at least)
Besides the screen mishap, there have been many small snafus that temporarily derailed us. Many, many components have needed resoldering or changes to the wiring due to loose connections. Parts like the 12 V voltage regulator and the power LED have needed replacement. One of the MOSFETs, which control the amount of power that goes to our heaters and fans, was especially curious. Sometimes it wouldn’t work, but then when I pushed on it just the right way with some random non-conductive object that was laying around, it would work again! We chalked it up to shoddy soldering (my specialty), but managed to fix it with some hot glue. Hot glue is like the duct tape of circuit boards.
By addressing and repairing these issues, we have shored up many of the most vulnerable parts of our incubator. With that out of the way, we could refocus on the primary function of our incubator: creating a warm environment for a baby. To date, we have now run 31 tests of different configurations of structural pieces, the mattress, and fans, and we have learned a lot about what doesn’t work. And a little bit about what actually does, too!
Shreya and I doing our best to provide positive reinforcement to our incubator in order to increase its heating performance.
There were many testing setups, but one of the most important things that we learned was that active movement of air is very, very important. My initial thought was the passive movement of air, as long as there was enough space to move, would result in hot air rising from below to warm the baby’s compartment. As it turns out, our design doesn’t work well at all when we only rely on this passive effect. Fan placement is absolutely critical to circulate the air in a controlled manner.
This setup shows the basic slit design we settled on to connect the heating compartment below to the baby compartment up top.
One of the most important developments we made in our testing was to have a fan pointing directly up through our slit, actively blowing hot air into the upper compartment.
As a result of the incremental changes and improvements we’ve implemented, we’ve reached a 7 degree Celsius increase in 20 minutes, from a starting temperature of 31.4 C to a final temperature of 38.7 C. There is still work to be done, certainly, especially with testing at lower starting temperatures, but we’ve proven that our setup can sufficiently produce the heat needed to keep a baby safe!
This is what our incubator looks like right now. Handles are conspicuously misisng, but will be added once the 3D printers you see in the background can be convinced to work.
At the suggestion of the design studio coordinator, Joel, we’re writing a local Tanzania patent for our device! This gives us the possibility to outsource manufacturing while maintaining the intellectual property. Likely, Shreya and I will finish drafting and submit it tomorrow.
—
Switching gears a little, I wanted to talk about a quick project I’m doing at DIT. I noticed that there were some missing floor signs in the 12 floor building where the design studio is, so I decided to replace. The following images chart the steps I took to design and fabricate these signs.
Fallen 6th Floor sign. This actually came in really handy because I took it to use as a reference for designing the replacements!
AutoCAD to create .DXF files for the laser cutter.
Laser cutting with Cleria, the expert!
Finished cut.
Peeled off the cut parts of the material covering.
Lots and lots of white spray paint!
Peeled off the rest of the covering for the finished sign!
Then, for what I thought would be the easy part: installation. Joel and I put up a couple of these signs with an epoxy glue that we were certain would stick, only to hear a discouraging crash of a fallen sign after only a few minutes! Tomorrow, we’ll give it another shot with copious amounts of double-sided tape.
—
Beyond those two projects, I’ve also been working with the Tourniquet Trainer, specifically with its blood pumping mechanism. It works with a simple hand siphon pump, but for some reason, the pump is hardly pumping water at all. In an attempt to improve the pump’s power, I made a one-way valve with materials I found around the design studio.
First iteration of my one-way valve. Fluid can’t pass easily from the top down because the ball gets in the way, but from the bottom up, the ball is stopped by the nail and there’s plenty of room for flow.
The biggest flaw in my first iteration was the ball, which was actually a glob of hot glue that I crudely tried to roll into a sphere. Luckily Joel took pity on me the next day and gifted me a ball from a ball bearing that just happened to be the perfect size! I installed it into my valve, which made it work considerably better. Sadly, though, the pump is still hardly showing signs of life. It may be defective.
This is much rounder than a hot glue glob.
In the last couple days we have in Tanzania, I hope to tie up some loose ends: finishing the patent, installing the floor signs, and making some useful documentation for the next set of Tanzania interns! I can’t believe how fast the time went, and while I feel that our projects are far from finished, I’m proud of the progress we were able to make and how much I learned along the way.
Miscellaneous Updates
Samaki Makange. Definitely the biggest fish I’ve eaten in one sitting!
The three of us wading to shore after our Mbudya Island trip.
Celebrating my 21st birthday with a brownie on a sizzling platter.
Crazy find at the Kariakoo Market, which is essentially an entire neighboorhood of street shops selling anything you can imagine. Who new I’d find what appears to be the Rice logo on the streets of Dar es Salaam?
Gorgeous view from the revolving Akemi Restaurant. I splurged for the 60,000 Tsh buffet (about $25), but I ate so so much that I think I made it worth it!
I’m clearly perplexed by what the Grand Restaurant called “Chicken Espetada”. We thought a more informative name was “Vertical Chicken”.