Friday, July 10, 2015

NSF I-Corps

Team "AutoEEG" - Me, Mentor Lou Bucelli, and
Grad Student / Entrepreneurial Lead Meysam Golmohammadi
This week I attended the NSF Innovation Corps (or I-Corps, for short) program. It was held in Chicago and attended by about 80 people. I-Corps is a program intended to teach researchers how to develop research creations for commercialization. NSF wants more of the science and engineering it funds to enter the marketplace, and so it is teaching academics how to be entrepreneurs. It follows the Steve Blank method, which as far as I can tell is similar or related to the Lean Startup approach.

The program instructors are all entrepreneurs. Each team includes three people: the Principal Investigator (the academic whose research lab birthed the technology), the Entrepreneurial Lead (typically a grad student or postdoc who will do the lions share of the business development), and a Mentor (an experienced entrepreneur who will guide the team). The program is seven weeks long. The kick-off week is held in person, followed by five weekly all-afternoon Webex meetings and lectures. Then we'll head back to Chicago on Week 7 for wrap-up meeting in person.

The main point of the program is to use the scientific method (data driven hypothesis testing) to discover who you customers are, what quantifiable value your product brings to them, and whether there is a viable pathway to deliver your technology to them. The trap academics fall into (we are told) is to have assumptions about what the customer wants. This leads many people to spend time solving a problem that customers might not care that much about, or that customers might not be able to use or integrate into their workflow, or pay for, etc. The only way to do this is to go out and talk to as many people as humanly possible to find out what their needs are, how they work, who makes financial decisions, etc etc. We are expected to interview at least 100 people at all levels of the customer pipeline over the next seven weeks. We've been given a pretty healthy budget to support us traveling around and meeting with people, but its still going to be hard (that's over 15 interviews per week!). The hypothesis-driven method makes a lot of sense. You make a hypothesis like "Neurologists at community hospitals are unhappy with how much time they spend reading EEGs". You can then test this hypothesis by speaking to neurologists, and depending on the outcome, set up another testable hypothesis. In this manner, you drill down until you understand which customer is feeling the worst pain that you can address (and that they can pay for).

The boot-camp was pretty brutal. For three days, all I had time to think about was how to advance our business concept. The set-up was similar to Shark Tank on TV. Our team would get up to discuss our technology and sometimes we'd get yelled at during our very first sentence. The honest was brutal, but also refreshing. The mantra was that we shouldn't be afraid to discover that our 'baby' is ugly :)

So I'm looking forward to the next seven weeks. We'll be busy, but hopefully we'll learn how to get a sustainable business launched...

Monday, June 22, 2015

Plane of the Ecliptic

This post has nothing to do with engineering, interfaces, brains, or Nicaragua, but I thought it was cool nonetheless. I took these photos tonight (6/22/2015, Philadelphia) and Friday evening (6/19/2015, Managua). You can see that the angle of the plane of the ecliptic is very different in the two pictures. I tried to measure the angle in both pictures (very rough estimate) and they seem to be about (62.5-28.3) = 34.2 degrees apart. This number should be roughly the same as their differences in latitude, which google tells me is (40d [Philly] - 12d [Managua]) = 28 degrees. That's pretty close considering my non-scientific measurement approach. Pretty neat!!!

Saturday, June 13, 2015

Nicaragua - Day 27

We finally reached the end of the program. Well, at least the first month, which is where my contribution ends. The last week was filled with a handful of interesting labs. One of them involved us picking through a big box of old medical equipment that someone had left in the school supply closet for us to look at. We took apart a blood pressure cuff, a nebulizer, and a few other odds and ends. The nebulizer was fun because it seem to follow the same basic pattern that all other suction and pressure devices follow. We tried to strip it down to its barest minimum, but were foiled by a lousy screw that stripped itself into smithereens and would not budge. In another lab, the students programmed a PIC to interface with a thermal probe to measure temperature and to set an alarm in case the ambient temperature was too low or too high. The students had to attempt to calibrate the system by mapping the arbitrary units from the analog to digital converter into degrees Celsius. The thermal sensor wasn't exactly the most accurate gizmo ever, so it was a challenging lab, but still enlightening. One highlight was the various methods students took to raise the temperature of their probe in order to calibrate it. One team put their probe in a ziplock bag and dipped it in the coffee pot only to discover that their ziplock bag was not exactly waterproof!

A thermal probe unit in a very leaky ziplock bag.

Yesterday, Friday, I took three students back to Nandaime hospital where we attempted wh tie up some loose ends. First thing on the agenda was the fetal doppler unit which I screwed up a couple weeks ago. Thanks to some friends of ours who flew in Thursday night from the United States, I had brand new mini USB Type B connectors in hand, and we were able to desolder the old connector and put in the new one without too much fanfare. unfortunately, I think we must have damaged some of the traces on the board during our first attempt to fix it and we were unable to get it working. I tried to guess where the broken traces might be so that I could short them but no luck. Well, at least we gave it everything we had. We also started working on an electric heater. We added a power cord (which somebody had cut off) but that didn't seem to be the only problem. We didn't have time to get to the bottom of it, but I'm pretty sure the heating coil needed a massive scrub because it was completely rusted through. For comic relief, we were handed a floor lamp that looked pretty new but we were told it didn't work. In trying to diagnose the problem, I happened to pull on one wire and the whole lamp snapped into life. I guess we got lucky! As the director of the hospital wryly noted, well at least you fix the lamp today. A bit of a hollow victory, but a victory none the less.

My biggest "engineering" victory of the week was replacing a light switch at my old homestay. You take your victories where you can!

The students are officially sufficiently trained to spend a month working in hospitals. They are a great bunch, and I expect they will all learn something and find some way of being of service to their various hospitals. I'm looking forward to keeping up with their exploits once I'm back home. And naturally, I take credit for all their successes and none of their failures :)

I'm off to the beach for a few days before heading back to Philadelphia. See you all soon.

Tuesday, June 9, 2015

Nicaragua - Day 23

Our third week in Nicaragua went really well. For the most part, students seem to be progressing well and making the most of their weekly hospital visits. In lab, we've been combining technical with non-technical experiences. One of our technical labs involved using our variable power supply to recharge a battery. Although successful, it was a little frustrating since the battery voltage does not vary much between discharged and charged. one of the non technical as we did was disgusting the nature of poverty and what it means to be a developing nation. We challenged are students and to define poverty and to discuss whether the existing metrics for being a developed nation or biased and if so how. We had some really nice conversations about whether poverty is an absolute or a relative benchmark, and the extent to which personal liberty and freedom are indicators of wealth. If you have money but you aren't free to spend it how you wish, then what's the point?

On Thursday we went back to the hospital in Nandaime. the first thing we tried to do was fix the microscope from the previous week which we thought only needed a new fuse. We install the new fuse and plugged it in and there was smoke and sparks! So we took the microscope apart again and took out the circuit board and discovered that it had been burned completely through. Charred to a crisp!
One seriously cooked microscope circuit board
In retrospect, I don't think that burning event was from our fuse attempt. I think that burning was a more significant event that happened when the gecko met its fiery end (see last week's post if you have no idea what I'm talking about). so that was a bummer, not much we can do there. on the plus side, we were able to fix several centrifuges and to put both autoclaves that we started working on back in service.
Fixing centrifuges is challenging because they typically contain three very stiff springs that must be wrangled back into place
We also cleaned and calibrated some more scales and even tried to fix a blender.

When you need parts, you must visit the local ferreteria, or hardware store.

On Friday we visited a central repair facility in Managua. There are a handful of engineers there who attempt to repair all the broken medical equipment from across hospitals in Nicaragua. some equipment is brought to their facility and some is fixed on site at the hospital. They are massively under staffed but they do some amazing work. The engineers are not formally trained, but rather learn in an apprentice style manner. But this system seems to work incredibly well, and everyone we interacted with could run circles around me with their eyes closed. They had a very well stocked machine shop, where they can engineer just about any widget of any size they need. they are also very creative in terms of rewiring and redesigning electronics is necessary.
An old school mercury thermostat that we spotted in a heater at the facility in Managua.
We showed them the picture of the fried microscope circuit board and they laughed and said that even if they had to redesign it from scratch with a reduced set of features, they would still get that microscope to turn on somehow. Super impressive. I'm not sure how much actual good we were able to do, although they gave us some work to keep us occupied. but we definitely learned a ton, and that's all that really matters.

So we have three lectures and three labs remaining, and then one more hospital visit on Friday. and that's it! The students will have officially finished their training and will move to hospitals across Nicaragua next week. I for one will be moving to a beach for 5 days where I intend to do very little engineering, and maybe try surfing again!

Wednesday, June 3, 2015

Nicaragua - Day 17

We are now officially past the halfway mark in our program, and the students are coming along really well. We are going through the trickiest of the labs, and the students are struggling and learning pretty effectively I think. Last Friday, I asked the students to take some time over the weekend to reflect on the first two weeks, and to think about what they still want or need to learn in the second half of the course before we unleash them into the wild. Judging from their weekend activities, I'm not sure how many of them took that message to heart, but at least I tried!

Our most recent labs included creating a variable DC power supply, and then using it to charge a rechargeable battery. The variable DC supply turned out to be fairly troublesome for many of the groups, since it involved a lot of soldering and a lot of components.  The DC supply consisted of a full wave rectifier, a handful of capacitors, a voltage regulator, and a potentiometer. A lot of students tried to compact their components tightly, but that often made for wiring headaches, especially since we were using untinned perfboard. To help out one of the teams, I rewired their board, being careful to arrange the components exactly as they were on the wiring diagram. This significantly simplified soldering job, as most of the nodes that needed to be soldered together we're all laying in a neat row. It occurred to me that simplifying the layout of the circuit elements is roughly equivalent to practicing good coding style in that you are massively reducing the likelihood of a mistake and making the overall design much easier for other people to follow. We had a nice classroom discussion about importance of good engineering practices in order to stack the odds in one's own favor for success.
Our variable power supply with circuit elements all laid out in a row, just like in the circuit diagram.

The resulting solder work is fairly easy to follow. You can see the ground bus going from left to right across the bottom.

For yesterday's lab, we used the variable supply to charge a battery. This was an interesting challenge, as students had to select both the power supply voltage and the current limiting resistor value. They had to make sure that they pick a power supply voltage that we could actually generate, a resistor value that we actually had, keep the power below what the resistor could safely dissipate, and keep the total charging time between 10 and 100 hours. That last constraint is because we were attempting to create a trickle charge. Overall the lab went well, but it was a little anti climactic, since battery voltages don't vary all that much between being charged and discharged, so there wasn't a whole lot to observe over the hour or so that we let our batteries charge up.

I'm looking forward to tomorrow and Friday, as it's back to the hospitals! Tomorrow we will return to Nandaime, and Friday we will be visiting a new facility in Managua which is run by the National Ministry of Health. Its not a hospital per se but rather a large central facility where they collect equipment from around the country to be repaired. Stay tuned for updates.

Friday, May 29, 2015

Nicaragua - Day 12

Hello again from Nicaragua! The weather has been getting steadily hotter and so our daily trips to the pool feel less like an indulgence and more like a necessity. On days that I teach, I typically spend my morning at home with my family. We spend the time playing Uno, reading, and going to the playground. In the afternoons while I go to class, they usually go to the pool or have some other type of adventure. On hospital days, my schedule is pretty different as I'm typically out of the house by 7 a.m. It's hard to believe that we are already halfway through the program. Two weeks down and two to go.

This week in class, we did lectures on a variety of equipment such as ventilators, oxygen concentrators, electrocardiograms, and defibrillators. We also did some pretty nice labs which get the students accustomed to dealing with hands-on electrical circuits. We did one lab where students had to construct a flashlight from a battery, a switch, and an LED.
a basic flashlight!
Not all of our students are electrical engineering majors, so this can actually be very helpful. We constructed a power supply using a transformer, a full wave rectifier, and a capacitor. We also got to blow a fuse with a battery, just to see what what happened. The students are asking lots of really good questions, and as always I'm probably learning more than they are in the process.

Yesterday I took 12 students back to the hospital in Nandaime. This is a small hospital in a village which doesn't have an engineering staff. This means that when things get broken or fall out of calibration, there is really no one to help them. Therefore, they had a huge inventory of things for us to look at. These ranged from the somewhat mundane, like a floor lamp, to tediously complicated, like an autoclave.
the control panel on the autoclave is simply a metal rod with some screws in it that open and close a series of switches when it is turned.
We successfully recalibrated about eight standard scales, and also repainted them because they were pretty grungy. We fixed 3 or 4 blood pressure cuffs which weren't working for various reasons. A couple just had some minor tubing problems that weren't too hard to fix. We were given a nebulizer which basically worked OK, except it was chock full of dust, which was taxing the motor. We clean it thoroughly, and returned it to the floor. We re-tackled the two autoclaves that we started working on last week. We took them all apart and washed out all of the copper tubing with vinegar, hoping to dissolve any mineral deposits inside. Even though these things are frighteningly complicated, we succeeded in putting them back together properly, which was a victory in and of itself! One of them appears to have been completely fixed! We still need to do a little testing, but it successfully pumped water into the chamber, heated it up, and then pumped it out afterwards. The biggest problem seems to have been that someone mis-wired the power supply switch. The other autoclave still doesn't work quite right, but it gets much further in its cycle than it used to, so I guess that its own little victory for now. Hopefully we will get another crack at both of those next week.

In one of the more bizarre things that I have ever seen, one of our students was taking apart a microscope that wouldn't turn on. She discovered that a lizard had crawled into the microscope and shorted out the power supply. The fuse had successfully blown to protect the electronics, but the lizard didn't fare so well and was found petrified on the circuit board!
a petrified lizard!
As if that wasn't enough, we found a clutch of lizard eggs in the microscope, and then we accidentally discovered that one of those eggs still contained a live baby.
a clutch of lizard eggs found deep inside the microscope
Unfortunately, we accidentally cracked that egg open, so the baby was born right in front of us, albeit somewhat prematurely. If you aren't squeamish, be sure to check out the video. Anyways, that was a first for me. Good luck seeing that in the university classroom!

In a more humbling moment, we were handed a handheld ultrasonic fetal heartbeat monitor. The doctor told us that it worked but gave noisy measurements and that the measured heart rates were definitely off. She told us to be careful since it was their only working one. The problem seemed pretty clear in that the connector between the ultrasonic wand and the base unit were hopelessly frayed and degraded.
the frayed ultrasound wand cable
We tried to re-shield the cable from the ultrasonic wand using some tin foil. Then I decided that we needed to replace the connector where the wand plugged into the base circuit board. After much effort, we were able to desolder and remove the connector.
The USB style connector was in pretty bad shape. At least two pins were probably not consistently making contact.
Unfortunately, after visiting about six or seven different shops, we were unable to buy a replacement connector. Finally I decided it was time to put the original connector back into the circuit board. However, no matter what I tried, the connector wouldn't go back in cleanly. After an incredibly aggravating couple of hours, it was time for us to leave, and I had to accept the fact that the connector wasn't going back into circuit board properly. We tried turning the device on, but it no longer recognized the wand. With my tail between my legs, we went to hand it back to the doctor. It's hard to tell whether she was really upset or not, but she thanked us for trying and sent us on our way. I feel pretty guilty that we ruined their sensor, and now I'm consumed with trying to get them new one, which will apparently involve a trip to Managua at some point. In hindsight, I should have taken a much more conservative approach with this instrument since it was their only working one. I think my hubris got the better of me, and I really only considered the best case scenario. I will try to communicate this lesson to my students today. I think it can be instructive when students see their instructor fail, and try to learn from that failure. Perhaps some good will come of this.

Anyways, despite that, it has been a fairly successful few days, and everyone seems to be learning and having fun. Hopefully next week we will get to visit a facility in Managua where they collect broken medical equipment from across the entire country. I am really looking forward to that!
Our hard work was rewarded with a nice pile of coconuts which we had to learn to open ourselves with the hospital's machete!

Tuesday, May 26, 2015

Medical Research Ethics

This op-ed appeared in today's NY Times:

Pretty interesting. Academics like to see ourselves as being trustworthy, upstanding people. But I guess is that the reality is we're just human in that we're corruptible by money, power, and pressure, just like everyone else...