Sometimes failure is an option. Actually, failure is often not an option but rather a necessary evolutionary step within a project. Learning from failure is how one eventually succeeds.
Allow me to explain….
You may be aware of our HURRB project. The goal of this project is to send a payload, via weather balloon, to the edge of space from inside the eye of a hurricane. The resulting video and GPS data that we could gather will probably be nothing short of extraordinary.
We tested the project in late May near Buffalo, Texas with near perfect results. The payload, made of a $2.36 cooler, had four GoPro HD cameras attached to its outside. On the inside was a collection of tracking equipment, including the first Android phone – the HTC G1. It would send its GPS position over the cellular network as long as it was below a few thousand feet. We also had a SPOT satellite based GPS beacon that would work from the ground to around 59,000 feet. We launched our balloon and it went to at least 94,000 feet and probably higher than that. We do not know for sure since we did not have a GPS tracker that worked above the consumer standard 59,000 feet mark. I estimate 94,000 feet based on the ascent rate and the amount of time the video shows before the balloon burst. In any case, everything went about as well as could be expected considering it was our first attempt at this project.
Then, in mid-June, Chris Erickson, a producer from CNN International Newsource, got in touch with me asking for some advice on a similar project he was putting together. He was not wanting to launch a balloon in to a hurricane, but rather was working on a story about how pretty much anyone with a little ingenuity could pull off this growing hobby of launching high-altitude balloons to capture stunning HD video from the edge of space.
I agreed to help out and actually asked if we could test some additional equipment on the CNN payload. We wanted to test out an APRS beacon, used by amateur radio operators to send out data and other messages via the amateur radio band. The GPS chip in the unit that we had purchased would work well above the 59,000 feet altitude mark and this meant we could track it to 100,000 feet or higher – which was Chris’s goal. I also contributed our SPOT satellite tracker and our G1 phone to help make sure we could find the payload once it fell back to Earth.
On June 23, exactly a month after we first tested HURRB in Texas, I met up with CNN, actually several producers and reporters from their International Newsource group, in Peachtree City, Georgia. I was joined by HurricaneTrack.com supporter and HAM radio operator, Kerry Mallory. It was his funding that provided the APRS unit that we would test during the CNN launch. Since every cool science project needs a nick-name, this was one was called “Yeager” after Chuck Yeager who first broke the sound barrier back in 1947.
With the help of the National Weather Service office in Peachtree City, about 25 of us gathered to set Yeager loose in to the warm Georgia sky near 8am ET, Saturday, June 23. The balloon and payload, which had three GoPro cams attached, took off almost straight up and stayed nearly above our heads for at least 20 minutes. We tracked it using the APRS website on our iPhones and were all very excited about getting Yeager in to the air. Now it was time for a quick bite to eat before we would set out to retrieve the payload after the balloon burst upon reaching its maximum altitude – hopefully 100,000 feet or higher.
As it turned out, something went wrong. The payload only made it to 66,000 feet before falling back to the ground. There was considerable disappointment and I wondered immediately what had happened.
The SPOT locator told us where to look as it sent out a signal every 10 minutes to a special page on our account with their website. There was no data from the G1 phone and the last APRS reading showed Yeager was at 5100 feet. It is important to note that we had the APRS set to beacon its location every two minutes. So it was entirely possible that the payload was on the ground between updates. And, since this system operates via the amateur radio network, it would be difficult to get a signal out to the Internet via repeaters unless one was very close by to Yeager’s location on the ground. In other words, the higher up the APRS unit is, the more repeaters “hear” its message. When it’s on the ground, it would have to be very close to either a hand held radio receiver or a ground based repeater.
We drove to the location of the SPOT tracker and found something shocking. Only the balloon, which had not completely shattered when it burst, the parachute and the wooden support dowels used to attach the balloon to the payload via thin rope were left. Attached to one of the dowels was the SPOT locator device – but there was no cooler, no payload. We were dumbfounded. Had someone stolen the payload? Anyone who knew where to look on the APRS website could track Yeager but only we knew the coordinates that the SPOT beacon was sending out. What happened? Where was the payload and the GoPro cams?
We looked all over the surrounding wooded area where the remnants of the balloon and parachute were found. No clues, none at all. We went up to the fire station which is along the Flint River and Hwy 16 between Senoia and Griffin. There we talked to local law enforcement and asked if they had heard anything. Who knows? Maybe someone saw what happened. The two deputies were very helpful and curious about the project. They made some calls and rallied up a couple of more deputies to put the word out about the missing payload. We searched a wooded area along the last known trajectory from the SPOT locator and the APRS unit but to no avail. The CNN crew needed to get back to Atlanta, sadly, without a conclusion to their story. At least not the conclusion they were hoping for. Kerry and I remained to keep looking, as did a tree climbing expert who had come along in case Yeager was high up in a tree top.
Then, something beyond all of our imaginations happened.
One of our support team members in Nevada, Paul Bowman (it was Paul’s G1 that we were using, and he too is an active HAM operator) called and said that the G1 had “phoned home”. This meant that the G1 had been turned on and it sent a message to Paul telling him where it was. We could not believe it. This meant, perhaps, that the payload was there too. Paul said, “It’s in someone’s house over on Lakeview Drive!”. I showed the deputies the coordinates and they said that it was just around the corner from where we were. So away we went to check it out.
Within a few minutes we were all parked in the drive way of Justin Garrett. The two deputies asked him about the phone and whether or not he had seen the payload. I approached and told him it was a black cellphone. He said, “Yes! I found a cell phone in my back yard a little while ago”. I nearly passed out. “Really?!??!” I responded. “Where is it?”
He showed it to me, it was on his work table in his garage, right where Paul said the GPS coordinates indicated it would be. He told me he was mowing the back yard when he looked down and spotted the phone. He picked it up and put it in his pocket, wondering who had been in his yard with an old camera phone. After he completed his chore, he went in to the garage and turned on the phone. This triggered it to send the message to Paul and within 5 minutes, a group of people he had never met before shows up in his drive way. It was incredible.
We told him what we were doing and asked if we could look around his yard. “No problem,” he said. After about 15 minutes of looking all over his large back yard, we we still empty handed. No payload. Only this phone that had somehow landed in his yard. This meant that it had fallen out of the payload and landed there. This also meant that the payload was probably broken in to pieces. We were closer to finding it, I thought, but this was a challenge. Where was the rest of the payload and specifically, the APRS unit. It had to be near by.
As the hot Georgia sun beat down on us, we became increasingly frustrated. Where could this darn thing be? Surely it had to have fallen near the G1 phone. Then it dawned on me. Ask Kerry to turn on his hand held radio to see if the APRS was close by. It just might be still transmitting its location and if so, Kerry’s radio may pick it up. It was worth a shot. My hunch turned out to be correct. Within a few minutes, we picked up the beacon from the APRS unit. It had to be within a mile or less of where we were. Since no other APRS sites could “hear it”, we knew we were close and it was on the ground or in a tree near by.
This boosted our spirits and we searched in the thick woods to the east of the Garrett’s house. He and his family even gave us a guided tour in to some trails to get deeper in to the woods, eventually reaching the Flint River. Ticks were attacking us like zombies hungry for flesh. It was hot and we were all quite miserable. Still no luck.
Kerry and I looked at his radio and scrolled through the data received from the APRS unit. Sure enough, it gave us the biggest clue yet. We had the GPS coordinates. They read: 33 14.15 by 84 26.03. I ran to the Tahoe where my laptop was connected to the Internet to convert the lat/long pairs in to something that I put in to Google Earth. I did not realize, probably because of the heat and the excitement of the day, that no conversion was needed. Still, I used some website’s conversion tool to convert what I had to decimal degrees. I did not realize that I had already had that info. The resulting output erroneously put the APRS unit, and presumably Yeager, really close to the Flint River but on its east side.
Kerry and I drove around to the edge of an enormous field. It was bordered to the west by a foreboding line of trees which surrounded the Flint River and its many smaller streams and tributaries. I told Kerry that the APRS location was only 1800 feet in. It was the toughest 1800 feet ever. We had very little with us to cut through the thick brush, briars and bramble that stood between us and the APRS and Yeager. Snakes, ticks and spiders were in ample supply. On we went, pushing deeper in to the woods. We arrived at the coordinates and found nothing. The trees were so thick that it seemed like dusk was falling, yet it was only 4:30 in the afternoon. The heat, even in the shade of the woods, was crushing. Despair set in and we had to give up. Let’s go back to the hotel and regroup was my thought.
We went by the Garrett’s to tell them of the defeat and called it a day. Maybe we can try again some time when we have more time and energy. We really wanted to find the payload if only to see what clues the video would yield. We had found the SPOT locator with some remnant pieces and of course the mysterious location of the G1 phone in the Garrett’s back yard. Yeager was somewhere in between, it had to be.
A week later we tried again with our search. This time, we had better lat/long info from the APRS unit to work with. Since I discovered from Paul that conversion was not necessary, we realized that we were looking in the wrong place the weekend before. We’re only talking about 300 feet or so, but in those thick woods, 300 feet made all the difference. Paul was sure that the 33 14.15 by 84 26.03 was in fact where the APRS unit was. If we found it, we would probably find the rest of the payload.
So once again, Kerry and I hiked in to the rugged Georgia woods. This time we came armed with tools to cut through the walls of brush that hindered us last time. However, the heat wave that had grown to historic levels was a real problem. We had plenty of water but the heat was overwhelming. We began fairly early in the day with our search. After only about an hour or so, we reached the exact lat/long pair and looked for the wayward APRS unit and payload. Nothing. Not one shred of evidence as to where it could be. We sat down along the banks of the Flint River to get a drink of our water and I remarked that it “should be right here!”. We were standing right where the coordinates said it should be. Anyone who knows Geocaching knows how hard it can be to find a hidden cache whose only clue is usually a lat/long pairing.
We searched for about 3 hours and had to give up yet again due to the dangerous heat that was smothering us. The forest was just too thick. If the payload had in fact landed in a tree, we would never see it. I told Kerry that maybe after the winter sets in and a good freeze comes, that we can come back a third and final time to see if the canopy opens up enough to let us look in the tree tops. We agreed on that plan for sometime in December and left the woods in utter defeat.
Time marched on. The hurricane season heated up and we had a small chance to launch HURRB during Isaac but decided against it due to the eye of Isaac being rather clouded over and parked over the swamps of Louisiana. If the Georgia woods was difficult to recover a payload, I was certainly not taking my chances with large reptiles and other creatures that nightmares are made of.
October ended with the historic landfall of Sandy and I kept busy with that event well in to November. Kerry and then began to plan our last attempt to search for Yeager. We settled on the weekend of December 15th. It could not get here soon enough.
After many hours of pouring over data, maps and talking about a strategy to find the payload, the day finally arrived to put it all to the test. I drove to Newnan, Georgia, not far from where we needed to look, last Saturday. Kerry drove up from Houston. We had a nice dinner and planned out our attack for Sunday.
I had contacted the Garretts to let them know we wanted to look one more time. We thought about coming in from the west along the last known “flight path” to see if we stumbled across the payload between their house and the Flint River. They had no problem with us using their house as a starting point. We hardly slept Saturday night knowing that we had a good chance of finding the lost payload the next day.
Sunday morning was chilly but not too cold. The sky was overcast and I thought that this was of benefit to us. I figured the bright orange tape that covered the bottom of the payload would stand out better if the sun were not bright, washing everything out.
We arrived at the Garrett’s house around 9:30 in the morning. We could immediately see a difference in the woods. The trees were all bare of their leaves. You could literally see through the forest now and up in to the tree tops. This was a good sign.
Kerry and I hiked in from the west and made our way to where we thought the payload may have landed, not far at all from the coordinates that the APRS had given back in June. Unfortunately, its batteries had run out long ago so we had only the recorded data to work with. We searched along a peninsula, part of a feature in the river called an “ox bow” with relative ease. No luck. There was no sign of the payload and we were sure of it. I suggested we go to the east side, right back to the coordinates again, and start fresh from there.
We hiked out within 30 minutes with a very light rain beginning to fall. This time there were no ticks, no snakes and few spiders to battle. It was so much better even though we were still empty handed.
We stopped in to talk with Justin about our plans of searching the east side again and on we went.
After a quick drive over to a dirt road that ran along the woods, we were ready. The rain picked up now to become a little aggravating but nothing we could not handle. It was better than ground temps of 110 back in June!
The hike up a logging road was incredible. The tall trees lined the road like walls. There was not anyone else out there. We saw no animals, not even many birds. It was still and quiet. Only the chilled rain drops broke the air with their pitter-patter on the decaying leaves that carpeted the logging road.
Then, we reached the point where we had to turn in to the dense woods and head west to the river. Kerry cut through using his machete’ and on we went. Even though the leaves had in fact fallen, the forest was still so dense that it became noticeably darker the further in we trekked. Still, not another soul was heard from or seen, save for the random shot-gun fire we could hear far in the distance.
I used my iPhone for GPS navigation and was trying to get us to the exact coordinates of the APRS unit. We stood there back in June but to no avail. For some reason, maybe the cloud cover, maybe the thick forest, but my phone was not working like it should. My position, a blue dot on the Google Maps, would drift even when I was stopped. This caused us to get somewhat lost. We were about 200 feet from where we needed to be, maybe more than that. The day was getting shorter with each passing second and we needed to get over to the river and plan our search strategy from there.
Both of us stopped and looked at our compasses and GPS devices. Mine finally settled down enough to point me west – towards the river. Kerry made it clear that we had to know precisely where we were before we moved. I paused and checked one more time. “West!” I said. “We need to head west and we’ll hit the river”.
I looked ahead and then it happened. As if to almost call out my name, I saw it. Through the woods, probably 200 feet away, I spotted the bright orange against an otherwise dull brown and gray backdrop of dormant trees and thicket. I told Kerry that I think I have found it. I pointed my walking stick like the barrel of a sniper’s rifle towards the object, barely visible but there none the less. He saw it too. We both pushed ahead, Kerry cutting a path through briars, saplings and other brush to make a bee-line to the orange object. I checked the GPS on my iPhone and sure enough we were headed for the exact coordinates of where the APRS reported its position back in June.
“This has to be it!” I yelled out. “What else would be out here that is bright orange and in that location?”
After another ten minutes of hacking our way through the woods, we reached the edge of the Flint River and an active tributary that veered off towards the northeast. At that moment, two grown men acted like little kids on Christmas Day upon realizing that the present of their dreams had been delivered by Santa. We found it. Yeager was about 20 feet up in some nasty, thick trees and vines almost completely intact. We were thrilled beyond belief and relieved at the same time. The APRS unit was almost dead on. Yeager was right where it should have been. In fact, we looked at that very spot in late June but could not see it due to the choking green vines, leaves and other growth that covered it entirely. My logic of waiting until December paid off.
Kerry hopped across the tributary, climbed up the bank on the other side and proceeded to cut Yeager down out of the tree canopy. It took about 30 minutes of some serious hacking away at the grip of the forest but it finally relented and gave Yeager up. He collected all of the parts, some of which came falling down in pieces like a pinata as he poked and prodded the cooler from below. Everything was there. All three GoPro cams, the APRS unit with its steel wire antenna still attached. We had succeeded after nearly six months of waiting.
The walk out of those woods was exhilarating. As if to protest our triumph, Nature threw one last curve-ball our way as the rain picked up to quite steady. We knew the Flint River would rise some after this rain and that had we not found Yeager just then, we may never have been able to reach it. The area has been in near drought conditions for several years and the fairly low river was about to swell, even if only a few feet, with abundant rain then and more to come later.
We reached Kerry’s truck and headed back to the Garrett’s house to tell them of the news. It was an exciting moment for us all. They were fascinated by the whole experience and were visibly happy for our success. I gave their 11 year old son one of the GoPro cams to use as his own. All three were perfectly preserved despite being exposed to the elements for almost 180 days. It was the least I could do for them considering the fact that we would not have known where to begin looking had Justin not found the G1 found and turned it on. Everything happened because of that one piece of insanely good luck.
Kerry and I went back to the hotel in Newnan and plugged in my laptop via HDMI cable to the widescreen TV in the room. Each 32 GB chip recorded almost 5 hours of video in stunning high definition. We watched each camera to finally see what happened and why things went so wrong back on launch day at the end of June.
To the best of our knowledge and from what we can gather from the video, the payload began to rock back and forth once it reached 50,000 feet and higher. Perhaps stronger than anticipated upper level winds were pushing on it, causing it to swing like a pendulum. With each successive push from the wind, the payload swung higher, eventually achieving a brief period of being weightless, like a child in a swing that is pushed to its limit. Then, the payload which weighed in excess of six pounds, would fall with gravity’s pull. At the same time, the giant helium-filled balloon is pulling in the opposite direction, trying to ascend. This went on for several minutes until there was so much give in the payload that it may have actually swung end over end at least once. Below is a video clip, slowed down to 12% of normal speed, that shows the rocking and subsequent break up of the payload.
After one swing too many, the payload jerked against the balloon just hard enough to rip loose the wooden support dowels used to attach the balloon to the payload. This is clearly seen on the video. The balloon then takes off, continuing to climb to burst altitude of around 100,000 feet. With the SPOT locator still attached, the balloon, the parachute, the wooden dowels and the SPOT sail away in to the dark sky above.
After a moment of turmoil, the cooler, mostly intact, begins its free fall from 66,000 feet. This is when something amazing took place. The G1 phone, which was zip-tied to one of the dowels inside the cooler, breaks free now that the dowels are torn away. It escapes the cooler and skips across the thin atmosphere like a stone skimming the surface of a lake. From there, it falls nearly 12 miles where it lands in the Garrett’s back yard….fully functional. The screen was not even cracked. The jolt did manage to turn the phone off and this is why when Justin powered it back on in his garage that it sent a signal to Paul in Nevada. It is hard to believe but the HTC G1 fell from 12 miles with only minor damage but still completely operational. The slide out keyboard was a little twisted but otherwise, the phone was not damaged.
As for the rest of the equipment? The three GoPro cams all survived their plummet as did the APRS unit. This brings me to my conclusion as to why our finding of the payload was so important.
One day, we will launch the HURRB payload in to the eye of a hurricane. It will be equipped with some of the best state-of-the-art video and GPS recording equipment on the planet. We will succeed in getting the payload up in to the eye of a hurricane, even if it takes years for the right one to come along.
Knowing that the equipment can not only handle the disaster that took place but that it can also be reliable in telling us where it landed is of paramount importance to our project. This was more than just about two guys who did not want to give up and needed an excuse to go in to the woods on a chilly, rainy December day. This was about proving that the science and technology work. That when the day comes that HURRB makes it out of the top of the eye of hurricane-X, that we know that when it lands again hours later, we will have a solid chance of finding it. Then we will get to witness something that no human has even seen before, not from these angles. Flying up the inside of the most powerful storm on Earth is a powerful motivator for us to make darn sure we can find the payload when the time comes. The fact that we finally found the CNN Yeager payload gives me hope that our HURRB project will one day make history.
This is why I say that, even in the face of failure, there is opportunity for success. We will learn from what went wrong and will utilize that knowledge as we move forward with the HURRB project. Then perhaps next hurricane season we will set HURRB free to embark on one of the coolest science projects ever attempted. I can’t wait to see how it turns out….
Have a blessed Christmas with your family and friends. Stay safe when traveling and I’ll have more here in the New Year.
Here is a time compressed video showing the launch up until just about when the payload began to break up. Questions about anything you read here? Send me an email: firstname.lastname@example.org