Edit 11/10. On second thoughts I am withdrawing this: Just a bit too complicated and confusing
Especially come to think it if seen from an Ardupilot partner point of view. I might propose something similar again as OpenSolo gathers steam, and if there is strong community and dev team interest. Cheers everyone especially Olliw!
Topic: Solo for Storm32 Gimbal controller developer.
Description: 3DR Solo with gimbal, purchased in the US and shipped to Germany. Stock is dwindling, and Solos are expemsive and difficult to get outside of the US. But e-bay still shows a number of new ones available in the $200-$250 range in the US. Add gimbal and shipping and cost should be in the $350 range,+/-$50
Planned amount $$ (USD): $300-$400
Estimated time for completion: NA
Rationale: Olliw has been doing some fantastic work with his development of the Storm32 gimbal controller, including integration work with Ardupilot and many contributions to ArduPilot as a whole. There are currently some issues with using StorM32 with a solo, see the “The Solo lives on” thread. This proposal would allow him to provide better support and enhance StorM32’s capabilities with Solo. It’s likely that Ardupilot as a whole would also benefit from this (beyond Solo specifics). See also this specific post, and some CANBus discussions.
I’d be happy to volunteer for procurement and shipping if this is accepted.
Can you explain how this benefits ArduPilot? I mean, I understand that making a peripheral work better with ArduPilot is desired, I just don’t understand how that is related to Solo.
Besides that, I find it quite dangerous opening the door to giving hardware (especially a RTF vehicle) to peripheral makers - what makes STorM32 different from other dozens of peripherals?
first, I feel honored that you consider STorM32 useful enough to even consider such a proposal
second, I agree with Francisco
isn’t it usually the other way around, that ArduPilot expects the peripheral makers to give hardware to them
third, I do not think that the main issues lie within STorM32 (i.e. any bug if there is any could be easily and quickly weeded out). I do think, or am actually convinced, that the main issues lie within ArduPilot and/or the Solo branch. So, the proposal is probably not directed at the right person.
btw, you also make a number of assumptions which may or may not be true
I do have btw expressed somewhere else my willingness to help within my limitations.
The Solo was unique in it’s day to have “smart shots” like cablecam hard to even match today. The 3DR gimbal used Mavlink to communicate the smart shot control. STorM32 optionally can communicate Mavlink and users have already used it to create new Solo’s transplanted to larger frames and using a gimbal that can use a bigger better camera. So the attraction is to get the capabilities of a smart shot machine with a better camera at a DIY price. The Solo main brd is being sold new on Ebay for $75, which includes a Pix2.0, HD video, WiFi rc control and telemetry.
With respect to your first question, you pretty much said it. It would benefit ArduPilot and the community by making it work better with Ardupilot. With respect to Solo, specifically: Solo is, especially with the efforts of the Open Solo initiative, a de facto Ardupilot RTF, with a large deployed base. And pretty much the only RTF at that functionality and price point. Edit: Just saw Mike’s response. Can’t further address the question better than he did!
StorM32 is quite unique:
It’s inexpensive, ~$25 (compare to AlexMos for instance. at 3-5x the price), especially for what it does, and uniquely integrated with Ardupilot.
It is very popular and enjoys a very strong following among the Ardupilot community. See for instance related threads on RCG, lots of interest and user/developer interactions. Close to one million views, over 10,000 posts on the main thread on RCG alone, for instance. (More views there, actually, than the main Pixhawk thread itself). And over 2500 posts and replies from the developer.
There is a strong need for it, and it fills a space in an area (aerial photography and video, gimbal control) fairly weak within the Ardupilot ecosystem. It extends Ardupilot possibilities, with its versatility allowing anyone to extend vehicles capabilities with gimballed cameras, from Mobius to Gopros to point and shoots to DSLRs. Applications besides pure photo and video include mapping, search and rescue, agriculture, etc …
It is uniquely positioned to leverage Solo’s SmartShots.
Storm32 is the result, as far as I know, of work by one single individual, who designed it, (both hardware and software), made it available at or close to cost, and who extensively supports it, for free. This is quite different from many other similar peripherals, offered with higher profit margins (nothing wrong with that, of course) and low or inexistent support.
In summary, I think we should encourage and assist such unique efforts, when possible and of course on a case by case basis.
Btw I haven’t talked to Olliw. I just saw that thread quoted in proposal, user issues and Olli’s response. For all I know he could decline if this was accepted! But I am somewhat confident I could take care of that.
I’d love to see http://github.com/OpenSolo working with more/other gimbals than the stock 3dr one. STorM32 sounds like it would be a great candidate. Pull request/s to the OpenSolo codebase to support this are encouraged !
I make available only the firmwares and hardware layouts, not the boards. I’m not involved in their production or selling and have not seen, and never asked for, a single dime from the sales (and my support is purely voluntarily)
I know this isn’t a lot of money, but wouldn’t time and resources maybe be better spent by developing an opensource gimbal solution?
There’s a massive gap in the ardupilot/diy/non-dji ecosystem for decent gimbals/cameras. Producing an opensource gimbal solution (if necessary with a simple reference hardware design) would allow hardware vendors to start producing boards and gimbals - there are probably dozens of hardware vendors that would jump at the opportunity and would foster innovation overall.
Fnoop: Agree about the gap. But a full solution is far away, if possible at all. Meanwhile this proposal would be a good first step in the right direction imo.
I should also probably add that this is not meant to be seen as an Ardupilot developer sponsorship. But instead a small gesture, no strings attached, that can make life easier for a developer with an extremely strong track record.
Are there not significant similarities with the components of a gimbal to the components of an autopilot? IMU sensors and attitude control, brushless motor control, encoders (rover), PID loops, EKF. The gimbal problems to solve are significantly less complex than that of an autopilot.
Fnoop, I misunderstood what you meant. I thought that when you were referring to a “gimbal solution” you meant a solution including mechanical gimbal (arms, motors, escs, etc …), not just controller.
If someone wants a gimbal to port smart shots into ArduPilot, great! That’s not what this is about, this is about making OpenSolo work with STorM32, so it is not connected to ArduPilot.
How so? Why is a Solo needed for that? The cheapest autopilot is enough to make it work better with ArduPilot - I think though that @olliw42 already has an ArduPilot autopilot.
I appreciate the list but none of those arguments make it unique. There are other inexpensive and/or popular peripherals for ArduPilot (besides me not thinking those are good arguments to give money to someone) and SmartShots should work with any gimbal (besides that not being an ArduPilot feature).
So, in summary, I don’t see anything that really makes STorM32 different from, for example, several different solutions for FrSky telemetry or GNSS or LiDAR or OSD. etc., etc.
Yes, when they want ArduPilot devs to make the driver for them. I’ll say though that I wouldn’t object if someone developed a peripheral and wanted an autopilot unit to integrate their hardware with ArduPilot - I still think that’s a very dangerous door to open though, so I wouldn’t support doing it.
This discussion is a little academic.
Even if OlliW42 would accept such a proposal. The major vendors like B&H Photo will not ship a Solo to Europe. They have an agreement with whomever bought the remain stock and sold it to them. Two, to ship a Solo sized package to Germany via the major couriers is $200-$300. The US Postal service will not ship any package to Germany with a Lipo because Germany will reject it at customs.
Yes, Ardupilot, Storm32 and Solo are all distinctly different, and separate. However, they all share a common attribute that give each greater utility to the end user when used together.
Solo, or artoo, arguably offer very useful features that have already proven to work well with Ardupilot. Features that go beyond the life and usefulness of the Solo air-frame. Will artoo, or variations of it not eventually be ported to more powerful boards with different wireless cards? We’ll have to wait and see.
After setting up a number of NT gimbals I can’t help but think how useful it would be to be able to plug a few NT motor/IMU modules up to one of my spare serials, or can bus, and easily implement a MAVLINK capable gimbal.
While OpenSolo, Storm32, and Solex may I add, may not be a core part of Ardupilot they go a long way to compliment it. IMHO I only see them helping to keep it current and relevant. While much of what is being asked isn’t Ardupilot specific, it is relevant enough to its use to start a discussion on the topic, if anything.