Fligh #1, the untold story
The 30th of May was a memorable day for all of us. After several months of planning, researching and developing our project, after countless hours of work, E-mails to the project list, conference calls, paperwork and authorizations, and 4 amazing weekend team meetings, we finally launched Spacebits 1. Our expectations were high, and so were our anxiety levels. We knew that we were ready to go, we did our best in testing out everything over the last weeks, but we also knew that we were dealing with an unpredictable system, and that was part of the fun too. In fact, what drove us from the beginning when we all started working on this mission was mainly that: fun. What a ride!
This is the very late report of the event, our first High Altitude Balloon launched in Portugal. What went good and bad, our conclusions and lessons learned and our plans for the future.
So here’s the untold story of the Spacebits 1 weekend, hope you enjoy it.
We went to Castro Verde on Friday and rented a few bungallows at the Castro Verde Camping Park (great place to visit, by the way), one for us to work on the project, and the others for family and friends.
By Saturday evening, everything looked really well, we had everything tested and working and most of the components were already assembled in the payload box, ready to go. Fernando Afonso and Filipe Varela went to Guia to buy a few L91 batteries and took the failsafe GSM/SMS module with them to field test it – it worked flawlessly.
For a minute there, it seemed as if we were going to have everything ready by the end of the night, then have a good night’s sleep and launch the thing early in the morning. Yeah right. We had a good long dinner at Castro’s restaurant, and then headed back to our bungalow, then we realized that the devil is in the details and all those small things missing took in fact hours to complete.
Celso and Fernando Afonso went to sleep at around 5AM and both Filipes stayed up all night and went for the launch with no sleep at all.
8AM. everybody’s up. Reality check with the team, the good news is that everything seems to be ready and working. Celso takes the car near the bungalow and starts loading everything: Helium bottle, AC inverter, balloons, parachute, cameras, electronics and other tools, and the Spacebits payload. The weather looks great (lucky us), the phone starts ringing as many of our friends are arriving and asking for directions. Some press and TV arrive too.
By 9h45AM we leave the camping park in 3 cars and by 10h15AM we are parking them at the launch site. Not bad. The launch time was 11AM and we still had some time to set everything up.
We unload all the gear into a big floor blanket to protect the equipment from dust and rocks. It’s imperative that the balloon goes up as immaculate as possible. Team starts the preparations.
It’s 11AM, it’s time. Everyone gathers for the big moment: friends, family, press and TV, about 30 people drop by. For us it’s a crowd. One of the place neighbors comes by in his car and asks what’s going on in extreme curiosity. We explain, he smiles.
João Pedro gave us a hand too by previously calculating the Helium quantities so he’s now filling the balloon himself.
Fernando Afonso and Filipe Varela check and get the payload box ready, turn the failsafe module on, place the batteries. Celso turns on both cameras and starts the CHDK scripts.
But something is wrong. Filipe Valpereiro looks awfully stressed at the car with his Macbook connected the main computer’s Arduino. The main computer, which includes all the sensors’ logging, the main radio and the cut off controller, isn’t working properly. Freaking Murphy!
The sun is well up, it’s a gorgeous day, no clouds, no wind, and it’s very hot. Several persons brought their sun umbrellas and beach chairs to the place, anticipating the wait. Pedro Pinheiro arrives in its Prius and kindly opens the truck to find a cooler box filled with ice, beverages and beer for everyone to chill (thanks mate!).
Time keeps flying, João Pedro keeps holding the balloon now half filled (we’ve spent ~4500L of Helium). Valpereiro comes to the launch spot, still very stressed, puts the main computer inside the payload box and starts debugging the thing. No luck. Something is terribly wrong. Celso joins him with a sun umbrella and a good dose of rationale to balance things. Not only the main radio isn’t working but the sensors’ logger has problems too. Team starts discussing plan B scenarios.
To make things worse, there’s no 3G coverage at the launch spot so we can’t update our twitter account or the website and there is this crowd of users waiting online for the live transmission without any clue, jumping to conclusions paired with the #fail hashtag. Bad.
It’s 11h50. Despite the good looks, the team is devastated with the last minute problems but wants to launch the HAB. Both cameras are working fine and the failsafe module has proven for its robustness so there’s good chance we can recover the probe. We decide to move on and launch Spacebits 1 without the main radio. Low profile was the key here. Most of the people weren’t aware of our problems.
By 12:00 Spacebits 1 takes off. Cameras shooting photos in unison, crowd claps the team. We’re all filled with an immense sense of mission accomplished and we can’t hide our pride and egos, our adrenaline peaked, there’s this photo of us looking at the sky that says it all.
In 10 minutes the balloon gets out of sight. Now that sense of accomplishment is replaced by one of uncertainty and anguish. For the next two hours we won’t be able to do nothing but wait for one single SMS. In case you don’t remember, the failsafe module is now our only means of communication with the probe but as it depends on the GSM network, it only works at low altitude and with mobile network coverage. The last SMS we had was from take off, at 822 meters, and then it shut.
We pack up everything and decide to head to Castro Verde and have lunch with friends and family. After wandering the village for more than one hour seeking for a place to eat, we finally settle for a small traditional barbeque restaurant and order some food. Celso takes the Macbook out of the backpack (finally there’s 3G), updates the @flyspacebits twitter account and apologizes the online community for the lack of the live feed. He also looks at the SMS daemon running on the server hoping for some new data, but nothing. He also configures the broadcast program with the cell phone numbers of the geocaching guys that dropped by. Everyone wants to be informed of the crash site when it occurs, if it does.
It’s now 2h00 past the launch time. We start speculating on chaotic scenarios. There have been reports of less than 2h HAB flights on the web. João Pedro pulls its spreadsheet file and recalculates everything for the 5th time. Fernando Afonso swears that he’d seen “something” falling from the skies on the horizon, just 30 minutes after the launch. Our hopes and dreams start to fade away. Celso starts thinking about the motivational #fail speech.
At this point Celso can’t take his eyes off the SMS log on the server. We didn’t sleep well, didn’t eat a proper breakfast and couldn’t have lunch due to anxiety and social pressure.
And suddenly it happens, Celso jumps out of his chair, literally, and shouts: “Guys, Stop! We have just received one SMS from Spacebits 1!!!”. In a rush, 6 or 7 persons run away from the restaurant leaving the others behind. Celso calls Fernando Afonso on the phone, he is having lunch at another restaurant, and says “Dude, we just got lat/lon coordinates in a SMS!!”, he replies “Really??? Awn man… Let’s go, let’s go!!”. Our geocaching friends had received the SMS too as we’ve put their cell phone numbers in the broadcast list.
And so the hunt race begins. Castro Verde hadn’t seen so much action in a long time. The probe had fallen 30km from the launch spot, to southwest, in the mountain near Ourique. At least 5 groups in 5 cars left in a hurry. Celso took his VW Touran with Valpereiro, Fernando Afonso, the camera man from “Elemento Indesejado” and the half filled helium bottle jumping all over the truck. João Pedro went with Filipe Valpereiro and Delfim Machado in another car.
Our navigation tools took us to unpaved roads of all sorts but finally, after some trial and error attempts, we reached the crash site. It’s in the middle of nowhere but it’s reachable by walking. To our surprise, Isa and Nuno were already there at the crash spot (these geocaching guys are pros at finding stuff), respectfully waiting for us. We held the payload box in our hands, it was in great shape, no major damage. The signaling LED lights weren’t blinking though, something went fishy inside with the crash.
This is the spot where Spacebits 1 landed, 30 km from its launch position
This is video (edited) of the Spacebits 1 flight. We have some photos too here.
Luckyiest people on earth
We found later, looking at the SMS logs, that we found the payload by miracle. The failsafe module only sent 4 SMS between just after he got network coverage during the last seconds of the fall and the crash event. When the payload crashed, several cables and electronic components got separated and the failsafe shut down and stopped sending coordinates. Had it been the case where the GSM card took more 30 seconds to sync with the network, or no network coverage was available (both cases likely to occur) and we probably would have never found Spacebits 1. Remember that we had no main radio.
Great footage found
When we returned to Castro Verde proudly holding the Spacebits 1 payload as a trophy, we carefully opened the box in extreme curiosity to find if the cameras had recorded anything. And yes they did, we found amazing videos and photos from above. Spacebits 1 was a success.
We found later that the cameras had shut before the flight’s peak altitude though. We estimate that the flight went up to 29km and the highest photo we have is about 22km. The reason was clear. They had shut before time because we stupidly wasted 1h of battery life in the ground while trying to fix the main computer and radio. Bummer.
No sensor’s data
Also, as expected, due to the main computer and radio problems we had, no sensors’ data was found in the microSD card on-board, including altitude, pressure, temperature, etc. Damn.
So here’s our balance of the Spacebits 1 flight:
What went good
- Failsafe module and client/server SMS software worked great and proved to be robust.
- Cameras scripts worked nicely. We may work on a few adjustments, most photos were overexposed for instance.
- Payload box
- Materials
- Suspension
- Parachute
- Balloon attachment
- Helium calculations and methods. We’ll share the formulas with you guys later.
What went bad
- Main radio and computer. DIdn’t work at the last minute. Had to be shut.
- Procedures
- Cameras batteries. Mental note: the cameras will be turned on only when everything’s ready.
- Helium filling too early. Mental note: Set the payload first, test everything, then fill up the balloon.
- Helium bottle out of the car. You don’t want to hunting with a huge helium bottle bumping in you car’s truck.
- Fill the car’s gas tank of the day before. Yes, we lost 15 minutes looking for a gas station.
- Cut off switch. Didn’t work because it depended on the main computer. We might want to rethink this. Yes, this means that Spacebits 1 fell to the ground with the pieces of the exploded balloon attached to the payload.
- 3G coverage on spot. Make sure you have 3G coverage at the spot where you’re launching the HAB otherwise you can’t use the Internet at all or update the website.
- Navigation software. Use a decent navigation software like TomTom to hunt the probe. Google Maps sucks at directions in remote areas, really.
Lessons learned
We learned a lot of lessons from Spacebits 1. Despite the normal and expected small improvements on robustness, software and electronics, we learned that we failed miserably at the procedures, especially the ones for the launch day. That’s something we’ll work harder for the next launches.
But there’s a lot of good news too. We learned a lot of theory with this project, we had to research a lot, read many articles and other HAB reports in order to prepare ourselves. We all learned a lot on electronics, materials, regulations, helium and gases, hacking and space.
All in all, now in retrospective Spacebits 1 was a huge success not because the main goal was accomplished but mainly because of the very important lessons we’ve learned from all the mistakes we did. We couldn’t easily learn about these errors on the Internet, you had to go through this invaluable experience first with all its risks.
Plans for the future
Boy do we have plans for the future. As you might expect, we’re completely hooked to this HAB thing. We’ll be launching other balloons in Portugal through 2010 and we’ll let you know about them soon. And before that there’s one important launch to be done: Spacebits 2. We need a stress free, low profile launch where the initial specs from Spacebits 1 actually work. We need to fix the main radio and computer problems, the procedure errors, and prove that the original plan works. And that’s what we’re about to do, real soon. So stay tuned everyone.
Thank you very much for your support. Special thanks to everyone who joined us on the 30th of May, you are all great friends. Thanks Isa, Delfim and Pedro for the great photos. See you soon.
