Uchusen Company

Uchusen is the world's first dronesat systems developer, based in London and Berlin. We design, develop and integrate minidrone and nanosat systems for spaceborne activities.

[uchusencompany]

Very cool article on The Atlantic re three US space startups developing low-cost satellite imagery leveraging the nanosatellite hw platform and the ISS. Great to see these budding businesses starting to take shape... and showing the way for other startups to follow in their wake.

(image credit: NASA)

[uchusencompany]

New year, new plans. Have started planning for the year ahead and the nodecopter project has really been bugging me for a while in terms of simplicity and extensibility... Looking at the libraries it seems it's more control- and fly-focused, but need to dig into this further.

Looking back at last year, the baseline x-drone developed was well and good for a simple, self-contained and relatively rounded solution, but not particularly sturdy in terms of failover. Vibration issues were really a headache and need to look into a number of issues there: erratic rotor behaviour, inefficiencies in the (handbuilt) vibration damping structure, overzealous (second-degree) PID coefficients, bullet connectors soldered as opposed to fixed etc. Although these issues are crucial for stable autonomous flight from A to B, issues of redundancy seem more likely to crowd out my time slots this year.

And hence the issue of autonomous formation flying, or more usefully for my purposes, autonomous drone failover. Have to look into how people like ETH Zurich have gone about it, but I'm more interested in a situation where the drones are physically connected to each other with a rigid structure and working as a team... More thinking needed as to how to put the rig  together and start experimenting! 

[uchusencompany]

At the UK Space Conference in July 2013, Tranquility Aerospace and Uchusen announced a novel collaboration project for 2013-2015 around Tranquility Aerospace's Devon One reusable rocket project. Uchusen's latest in-house engineering work this year was presented, including original nanosatellite sensor boards development, a fully integrated autonomous minidrone prototype, and associated flight data from multiple low-level flight tests.

As Tranquility Aerospace finalises its first engine prototype, Uchusen's role is to help develop high-precision flight control systems for the liquid rocket propulsion system, as well as nanosatellite sensor systems for future payload deployment into Low Earth Orbit (LEO). If all goes well, a first integrated hover test is currently planned for Summer 2014, followed with a trial launch in 2015.

About Tranquility Aerospace

Tranquility Aerospace Ltd. design, manufactures and assembly for the space industry and is currently developing its own launch vehicle, so as to provide a full turn-key solution to our customers. Website: www.tranquilityaerospace.com

About Uchusen

Uchusen is the world's first dronesat systems developer, based in London and Berlin. We design, develop and integrate minidrone and nanosat systems for spaceborne activities. Website: www.uchusencompany.com

[image credits: (all images) Uchusen Company; (excl. bottom-right image) Tranquility Aerospace Ltd]

[uchusencompany]

Only a few hours to go now... until the UK Space Conference 2013! #UKSC2013

[uchusencompany]

Getting ready for tomorrow's UK Space Conference 2013 in Glasgow...

[uchusencompany]

Wow, and to think this was just a year ago...

[uchusencompany]

Interesting... Japan's Axelspace, a spinoff venture from the University of Tokyo, plans to develop three 5Ux5Ux5U 60kg microsats by 2015 at total cost of approx $10m. We assume that this does not include launch, so the total bill is still unclear. More after the link (JP).

(image credits: Axelspace Corp)

[uchusencompany]

ESA is looking to crowdsource Parrot.AR drone docking data, hopes this will lead to autonomous ISS docking algorithms. Interesting idea for such a large organisation, and further proof that the family of technologies around drone development are the platform of choice for grassroots space R&D in the medium-term. 

[uchusencompany]

Finally printed Uchusen's first fully-functioning home-designed Arduino-powered PCB board for an upcoming aerospace project. Naturally can't give away too much at this time, but it's clearly a step forward for Uchusen. Just under a year since I started this blog and experimenting with various electronics components, all the hard work has finally come together... and I realise how much more is yet to come!

Now that I have the foundation and ironed out all the kinks in the process, it's time to work on a more fully-fledged system and get closer to the reality of a nano-spacecraft built from first principles.

[uchusencompany]

Casual note next to download link for pre-release version 2.0b7 of Processing software reads "475,382 bug fixes"... Well I hope that's a typo because that's a hell of a lot of bugs just to manage let alone fix!!

---

Note: Razor IMU code works fine in the beta build, even though Processing itself is a massive 167MB beast zipped up. Assuming that there's a bunch of beta bloatware in there that will be cleaned up when the build goes legit.

[uchusencompany]

According to the BBC, the MOD have just purchased 160 mini drone copters for recon and intel gathering. The 14g, 10cm-long package contains a mini-IMU, video feed, 30 minutes worth of flight and... well that's pretty much it. It's going to cost the taxpayer £125k a shot for one of these, so a rather expensive piece of machinery if it gets lost. Admittedly, the price includes maintenance and support, as well as the portable tablet/screen/radio combo, but not sure how much maintenance is going to be possible if one of these things is lost along the way. Nice how drone technology is coming down in size, not sure about the price or the privacy implications but I'm old-fashioned that way. 

[uchusencompany]

Well, didn't make it to Conrad but instead implemented a quick-and-dirty setup with Arduino Uno and couple of wires. Not ideal but will do for a first appraisal. Tested the sonar straight from pin#2 and output fine with stated 58us/cm. Was coming up with 60-odd cm which I thought strange, before I realised I was coding around the piece and my head was getting in the way of the pulse profile! Moved out of range and the ceiling came up pretty steady at 220cm, which is about right.

Merged with code for the temperature/barometer sensor and worked out fine outputting at 9.6kbps. BBC weather info gave 1015mb for local air pressure but the uncalibrated sensor was reading 1005mb-odd, giving an altitude discrepancy of 55m to 66m... need more testing to see what could be the difference or which reading to use. GPS unit not getting any readings for some reason with the original code, maybe cloud cover is too heavy (?) but L-band shouldn't be affected regardless. Weird. Try again tomorrow, hopefully better weather though I doubt it.

In other (not good) news, the solar power system is not responsive for some reason, so first assume the batteries are empty as haven't charged them up for a while. The Libelium Li-Ion battery charger (step-down) unit also didn't respond to backup USB plug-in power, so maybe this unit is dead too. Will try tomorrow in whatever little sunshine comes through the snow clouds... if the solar panel check LED still doesn't light up, will be needing a new charger unit.

[uchusencompany]

Ultrasonic sensor is here, a mere 3 days after purchase, and no customs tax / shipping either. Tinkersoup is just great - Berlin power! Definitely order from them again.

Only problem is that the screw at the back of the sensor is so big (I assume to compensate for the sensor on the front) that a regular breakaway header is too short to plug into the breadboard. Time to head over to Conrad tomorrow evening before they close and grab a couple of 90-deg breakaway headers, plug into Arduino and my solar power system, and see if the conversion from pulse width to distance really is as easy as it sounds! Also curious to see how stable the signal will fare both indoors and out, and if any temperature-related variance crops up, but that will have to wait for the weekend.

[uchusencompany]

Stanford Univ TiO2/Si-nanotech batteries - the future of nanosat battery packs? From the following article in MIT Technology Review this week, Professor of Materials Science and Engineering Cui suggests that new TiO2-encased sulfur-particulate batteries could boost battery life 5x, energy densities 2-3x, and boost cycles outside the 400-1200cycle range of Li-Ion batteries and closer to the 3000 cycle range required for electric car batteries. Although for now the tech is straining at the 1,000cycle pt, this is promising tech for commercialisation in the next 2-3yrs.

My particular interest is if this technology can help further battery life in nanosat power systems. The following Aero Eng Master's thesis from Univ of Liege two years ago focused on investigating thermal issues in the successful OUFTI-1 nanosat notes in p.12: "Batteries are notorious sources of trouble in nano-satellites and have even caused missions to fail".

Multiple reasons can be thought of, but the ones that come to mind for me are: low cycle lifetime (lead to loss of charging ability), poor power management and batt protection (overcharging that can lead to explosions/fire), mechanical deformation of batteries (see the picture in the thesis, to be remediated by mechanical plates), temperature performance below zero.

The article didn't have anything to say about the last two items, but if a prototype battery with this material can perform well at low temperatures (ie. -35 deg C < t < 5 deg C), this may eliminate the current need for a separate heating system to keep batts at a high enough temperature (currently > 5 deg C) to work at all in the cold vacuum of space. Maybe some combination between thin film lithium ion battery design (for mechanical issue mitigation) and this new material (for the rest) could eliminate both problems and usher in a solution for longer-term nanosat missions.

[uchusencompany]

Wow, Adafruit's Lady Ada named Entrepreneur of the Year by Entrepreneur Magazine! Great news and recognition for her hard work over the years... my first couple of kits were from there, regardless of the extra tax to be paid to ship over from the US. The implementations are a bit quirky, such as the Xbee implementation etc (see other articles on this blog about that one).

But, you could put the kit together with some time spent perusing her images and detailed steps/hints on the website, you had basic software provided that actually did the job out of the box without detailed knowledge of chips, signals and electronical engineering, and the units themselves were pretty much sturdy and worked! FAQs were there, useful and responsive forums, goods generally delivered fast unless stuck in customs (nothing much that can be done there).

Now that I'm working on more complex systems, I feel I have reached close to the limit of what Adafruit can offer me, but I have lost count of how many parts I have bought and run successfully for Uchusen prototypes on Arduino. For example the (fish-eye-lensed) TTL Serial JPEG Camera was a great one, with virtually no-one else offering the same with such usability, although soldering wiring direct to the PCB and not being able to break out of the one-pic-only config was not ideal.

To sum it up, quirky but solid. A boon to all those of us who are not born electronics geniuses, but still want to spend the time to learn and hack together something cool that could be useful in other fields of industry outside the hobbyist community.

[uchusencompany]

Less than ideal... JAXA admits spyware siphoned off details of its next-gen $48m/pop small satellite launcher Epsilon*. Considering this rocket reuses the successful HIIA rocket as first stage, and the successful M-V design for the other two, it is highly unlikely the spyware was looking for the program manager's credit card details or trying to help sideload ringtones. Rather more likely one of Japan's neighbours looking for tried-and-tested designs to help undermine the commercial feasibility of the product. Considering JAXA is spending $1/4bn upgrading this successful technology, that's one expensive data breach and a sharp reminder industrial sabotage in aerospace is still a modus operandi in East Asia.

*Note that "small sat" is relative: Epsilon is supposed to be launching 1.3T sats into orbit, so not exactly microsat/nanosat territory.

(image credits: JAXA Epsilon Launch Vehicle pamphlet PDF)

[uchusencompany]

Now that I'm back from a trip to Boston/NYC/London/Paris, time to look at building a more complex Arduino control system. As I have been increasingly looking to UAV development to help as the basis for the development of low-cost and responsive autonomous spacecraft and rocketry, I have come across these Hexacopter kits from 3DR and jDrones...

Doesn't seem to be much in terms of spec difference at least for my uses. It seems like the basic platform is there, but the real items that set this apart such as IR, Sonar, long-distance TX/RX still seem to be options which doesn't make much sense but still look like they can be bolted on. Comparing the two specs gives the following:

From jDrones, including IMU, flight control, non-uBlox GPS and magnetometers -

ArduCopter Hexa v1.0 - Fully featured UAV development platform.

1 x ArduCopter Hexa v1.0 24" Frame (28cm arms)

6 x Motors designed to ArduCopter use, 850kv (can be changed to bigger)

6 x ESCs designed to ArduCopter use, 20Amp

6 x Propellers designed to ArduCopter use, 10x45 (can be changed to bigger)

4 x HighRise 20cm landing gears

1 x ArduPilot Mega Flight controller board

1 x ArduPilot Mega OilPan IMU board

1 x Connectors/Cables between APM + IMU

1 x Set of 3 row pins for APM

1 x MediaTek 3329 GPS + 10 cm Cable

1 x HMC Magnetometer + 5 cm Cable

Everything else not listed here are optional and ArduCopter can be expanded with them of users can build their own expansions. Eg. this kit does not contain: Camera, Video systems, Batteries, Radio or other devices. Needs: Radio, Receiver and Battery to make it fly

From 3DR incl sonar bracket and uBlox but no sensors (roughly $35 for the sonar unit itself though not sure re range)

3DR ArduCopter Hexa-B - Fully featured UAV development platform.

ArduCopter Hexa 3DR KIT, Electronics Contains the following:

2 Fiberglass Main frame boards 

7 Aluminum arms, 5 Black arms and 2 blue arms

2 Fiberglass Carrier boards

Fiberglass Landing gear

Power distribution board, power cables and signal cables for PDB-APM

All hardware needed to assemble the components above

6 Motors (850 or 880Kv)

6 ESCs

6 Propellers (either 10x47 or 11x47)

Ardupilot Mega 2.5 kit unassembled with choice of connectors and GPS unit (regular MediaTek or uBlox)

This Hexacopter kit is based off the 3DR ArduCopter Quad. With two extra motors the lifting capacity is increased to over 1kg. The PDB still mounts neatly between the base plates but access to it is made easier by thumb nuts that easily unscrew to remove the stackup. This also lets you attach and remove APM with screws more easily.

Hmm... decisions, decisions!