Refund for large pile of Mesh devices


#1

I bought a number of devices ~15 I think to do prototyping and see if this new Mesh system will work for our needs. Unfortunitly they don’t work becuase it is not possible to even register them (how do tested products not even make it past step 1?). I wasted a lot of time on them and no one seems to be making progress solving the issue Here At what stage can we agree these products are not fit for purpose or faulty and refund the money.
They are all in their boxs gathering dust.
Has anyone sort a refund yet?


#2

I think @kubark42 might be interested in buying them off of you…


#3

Solving the Android issue is possibly the highest, or nearly highest, priority with the Engineering team (another being the IPV6 issue with certain routers, which is confirmed to be fixed (at least for me) in the upcoming rc.26). We are having daily squad meetings about mesh issues overall, and this always gets a lot of attention. Our Android developer is working (at least) full time on determining a fix.

I know that you’ve been very patient in waiting for your shiny new mesh devices to arrive – I was possibly less so! But I know we will get these things sorted out.

I don’t think it was well understood how much different Android versions/implementations differ under the hood. But moving forward, we will definitely be testing against a larger sampling of the various Android devices that are available.

I said this elsewhere recently, but what Particle is doing is really hard and to my knowledge we’re further along on delivering an enterprise-grade mesh ecosystem than anyone else. Are we there? No. Am I confident we’ll get there? Absolutely.


#4

I can certainly understand why those are very high priorities, but it is amazing how non-functional and incomplete these Mesh products are.

I’m testing with a very simple setup of an Argon and a Xenon, and the number of times the whole system locks up requiring all devices to be cold-reset is just crazy. And it’s just wildly unstable. Earlier today i was sending sensor data from the Xenon to display on an OLED on the Argon. About 30 minutes into my test, the Argon starting showing showing garbled data and then stopped updating. Based on the test data I was also echoing to the cloud the garbled data occurred for 60-90 seconds before the Xenon was knocked completely off the cloud, flashing cyan with a few yellow flashes every 30 seconds. I had to cycle the power on both devices to get it running again.

I’ve read these stability issues are another of your “high” priorities, but at this point it will be hard to ever trust this Mesh platform for non-toy applications. I couldn’t imagine an industrial setup relying on this. You delayed 5 months to get the kinks out, and I wouldn’t complain about the delay if the product were good, but at this point calling it half-baked would give you too much credit.

Not to mention all the things that are supposed to be in there but no where near “high” priority. Low power modes, high availability networks, external mesh antennae. Without those the product is really just not ready.


#5

I can assure you that stability is certainly something that is on our radar. As Dave mentioned, we do have daily squad meetings to discuss matters at hand with Mesh. We recognize that Mesh isn’t as fully featured as we would like it to be, and there have been some shortcomings with reliability. But I can also assure you that we are hard at work because everyone is determined for the vision of mesh to be stable and fully featured as you are anticipating it ought to be.

We are always monitoring our forums and working with our community members to diagnose the issues at hand, and the community’s input does matter. Reliability is a concern for many, and that is why that has been our first focus. Dave has mentioned that Bluetooth is a current top focus–and he’s right, it is a current top focus because we have been hard at work with other reliability issues such as general connection stability–of which rc26 is a result.

With that said, it isn’t to say we are completely done with general mesh reliability. It very much is and will continue to be a high priority for us. At this point, we have identified several bugs in mesh and have attempted to address them in rc26. Once we release it, we will continue the feedback loop from our community to see what has improved and what still needs attention.

As I said, the community has spoken–and input such as yours has defined what gets fixed first. The community’s concern with your Xenon nodes is a legitimate one, and that’s why we have prioritized it be addressed in rc26.

In terms of mesh feeling “incomplete”, due to some reliability concerns, not all features of mesh are fully available just yet. For instance, we have opted not to fully implement high availability networks just yet in order to eliminate extra variables when diagnosing and addressing reliability. So while it is not all visible on the surface, there has been a lot of thoughtfulness put into the infrastructure of mesh from the ground up to accommodate such features. Just because they aren’t visible or available yet doesn’t mean that no work has gone into them. There’s a lot of “behind the scenes”, if you will.

With that said, once we are in a more comfortable state, we will prioritize on adding these other functionalities. Although it won’t be as simple as flipping a switch, we won’t be working from the ground up.

You say that you wouldn’t ever trust Mesh in its current state for “non-toy” applications. In its present state, I can certainly understand that. But I can assure you that we have a determined team at work whose vision is to get it to a state that you would find it production ready.


#6

Just wanted to mention that one is supported (partially) already:


#7

Thank you, good reply. Most people buying your products are pretty technical so one of the big frustrations is the lack of information. For example some people are able to setup these devices and some are not, so as you narrow it down please keep us in the loop and share any work arounds you have.
I still don’t know if it is device specific, Android version Specific or Location based etc. Most are willing to help if kept in the loop with testing etc.
Why not have someone provide a daily update, harness the power of the community.


#8

Hi Toogooda,

Just for the record I’m no IT expert, I am in fact an old fashioned Mechanical Engineer, so working with computing baised systems is something out of my comfort zone. That said I’ve been of the edge of the IOT world baised in London UK, and been experimenting with Partical products since the Core.

I understand your frustrations with the Mesh, and totally understand the disapointment that has resulted in the setup requirements of using BLE to configure the devices before you can even ‘work’ with the platform. I saw this coming, about a week before my delivery arrived, when I was ‘reading up’ in expectation of my delivery. I know all my bluetooth devices are not going to play the game. I’ve worked with Bluetooth before, never succesfully and for me I’m waiting until a USB configuration option becomes available. So my devices are still brand new, in their boxes, on the shelf and I’m going to wait until that gets sorted out before I have a play. Also I know there are other higher priority issues the particle team are working on before they address that issue. I’d love the team to address the set up issue first, becasue until there is a relaible setup option fixing the working bugs will be harder becasue many people will not know if there problems are functional issues or setup issues. But I’ll wait.

Bluetooth is actually really painful to develop with for a whole host of reasons. I’ve even pleaded with representatives of the ‘Bluethooth special Intrest group’ to assist small scale developers, but at the time they were not interested. Bluetooth has had a massive uptake by manufactures, and if your Apple or Samsung with considerable resources then you get access to development resources the like of you and I never see. I can express opinions about Bluetooth till the cows come home, but thats for another day.

This mesh idea is pretty new, its been floated around as an idea for quite a while, and I even remember discussing it with developers for long range low energy radio systems when the whitespace spectum opened up here in the UK, but to my knowlage little ever came of it, although I’m happy to be corrected on that.

So with all that said, I’m super excited to see a mesh enviroment come to the fore.

This really is at the bleeding edge. Trust me experianced developers and enthusists such as ourselves are very much in the same boat, and there will be many tears before bedtime before things mature. The community is here, and we are using the forums to share frustrations and team are monitoring and working very hard solve many of the issues. From the user perspective like us it is worth monitoring other peoples problems and looking for the diamons in the coal that give us the spark of inspiration to solve problems or realise we are not alone in the mix, and when that happens sometimes we need to wait until the solutions are found.

I’m sure if you want out of ‘the process’ someone will happily purchase your devices. But I hope you find the strenght to persevere. It’s OK to gather them all up, put them in a shoe box and leave them there for a few weeks and come back to them later, when you can be bothered. (God knows I’ve done that myself). I even said out loud, wagging my finger “YOUR NOT COMING OUT UNTIL YOU PLAY NICELY” at a shoe box. And put them on the shelf.

But even with all that said, I don’t think you mentioned what made you want to play with the Mesh in the first place, what was the idea that made you wait this long to get your hands on them?

Anyway have heart, I hope you don’t give up. But yes it is frustrating. I know this is VERY possible, but it is difficult, really difficult. The community here made up of Particles own people and many many unpaid, enthusists are phenomenal people, and this is one of the strongest community groups I had the pleasure witnessing.

“The road is long, sometimes it lonely, sometimes its stoney and sometimes it passes though barron lands. Its ahead of you, and can only be walked one step at a time. But you will meet interesting people and see new lands along the way.”

Liam


#9

@VideoLiam :100::+1: for that!


#10

many android app devs have support for ~5 to 20 handsets and outside of those generally do not assure that the app will perform as expected. bluetooth is a known minefield. i wish the android app dev at particle the best.


#11

I can certainly understand those frustrations too. From my past experiences of community management, one of the hardest things to do is be able to effectively communicate to the community at large. Of those having problems, everyone is going to be experiencing something somewhat different. Quite a few will fit a specific issue, but there will be those with outliers. Still more, there can be instances where two community members may be experiencing the same issue, but exhibit different symptoms because of variables (such as using Android vs iOS).

We have had some discussions on what we think would be best to communicate to the community. Our current methodology is updating our pinned topic at the top of the community pages here. Looking it over, this topic is reasonably up to date with our latest discoveries and developments. For instance, the reliability issues you had concern with around the Xenon nodes on your mesh gateway–we believe the major factor behind this was actually a bug with OpenThread (You can find this under Issue 4 on Device Issues on our topic).

If you think it would help, in addition to updating the topic–it would be beneficial for us to make a post each time we change to the pinned topic, I can certainly relay that back during our next meeting.

In terms of your concern about not knowing what the issue is behind mobile device setup and what’s causing it–to be completely honest, we’re still not 100% confident either. We have some theories, and actively investigate them as they come up–so much so, that I’ve heard the proposed cause radically shift from one daily meeting to the next. When it is changing so wildly, it’s hard for us to communicate it when we aren’t even able to pin down anything internally.

To try and give you a snapshot, my current understanding of the BLE setup flow is primarily that iOS seems to be generally more reliable at this point. But I can’t guarantee that iOS will work without issue–which is why we want to be reserved on how we communicate this. I think it would be even more frustrating for our community at large to be told that any iOS device will do–and after they go out of their way to get one–it, too, doesn’t work for them. For your specific circumstances, I would certainly recommend you try an iOS device if one is easily accessible, but I don’t want you to go out of your way for it and expect it to work.

With Android, we have had theories of it being an Android manufacturer model, an Android Version, and even a BT radio manufacturer housed within certain Android phone models. It tends to change relatively often though, based on community feedback.

For instance, we had a day where we had theorized the issue to be around Android 7 on a certain subset of manufacturer models. Not even an hour after the meeting, I had handled a customer ticket who tested his setup with 3 different Android phones–a Samsung series, an LG Series, and an Huawei series. Each one had a different version of Android on it (6,7, or 8).It’s feedback like that that puts us back to the drawing board and sometimes makes us hesitant to communicate that out when it would get dialed back so fast or become seemingly irrelevant.

The goal isn’t to avoid communication or hide anything–but we also don’t want to give false positives or also steer the community in the wrong direction with potential causes.The feedback is immensely valuable from our community, and we want to make sure we are driving it in the right direction so we don’t allocate our community’s time and resources on something we aren’t confident in.

With that said, this is also in part why we took our time with rc26 and want to ensure a level of reliability so that our community can try it and engage with us, to figure out what still needs work and where we’re making progress.

Personally, I’m actually pretty excited for our rc26 release. :slight_smile: If you’re able to get it setup (perhaps with an iOS device, if one is easily accessible for you and you’re up for giving it a go), I would love to hear your feedback!


#12

@VideoLiam thanks for the background on the bluethooth development, I may well have to put off plans and look at other options for now.

I don’t think you mentioned what made you want to play with the Mesh in the first place, what was the idea that made you wait this long to get your hands on them?

For comercial reasons I can’t go into detail but we are working on sensing and automation in environments that will include very remote sometimes off the grib networks. We have working protoypes using ESP8266/32 and RF with directional recievers and MQTT back end. We are not happy with standard TCP as its far more than we need and slow at connecting and we never move large data. We are investigating alternatives for the local hives and these devices advertised some of the features we are looking for. I am getting good results from ESPNOW but using an ‘already working’ Mesh solution it what attracted me to this.


#13

Thanks @mstanley great reply,

Yes I do , I have looked at this post and I think the sooner people are directed to it the better so we can see things are boing done. Our industry is legendary for superhero level commitment to solve somthing but getting no credit due to poor communication. The posts look like they are days old becuase the old posts are getting edited.

For instance, we had a day where we had theorized the issue to be around Android 7 on a certain subset of manufacturer models. Not even an hour after the meeting, I had handled a customer ticket who tested his setup with 3 different Android phones–a Samsung series, an LG Series, and an Huawei series. Each one had a different version of Android on it (6,7, or 8).It’s feedback like that that puts us back to the drawing board and sometimes makes us hesitant to communicate that out when it would get dialed back so fast or become seemingly irrelevant.

This is actually exactly what I am looking for “We are investigatiing to see if it is …but can’t garentee … will sort it for you” It makes me think how can I help. As long as you are clear like you have been in this post it is testing opertunities not false possitives.

What is it? New Android App? I am happy to beta test if you need it.


#14

Hi toogooda,

I understand the issues around comercial sensitivity. From what you have said I can see what attracted you to the mesh concept. I’m not a 100% convinced that BLT solution will have the required range, but it will certainly have engergy requirements that mesh offers and deployment scaleability. Maybe Lora will have a better range. But lets not reinvent this wheel. Maybe its an option for Mesh II the sequale…

At the moment there are very few scalable mesh protocals that I know of (but I am just one man in a very fast moving technology space) so there definitly is value to tackling the mesh side of the problem. To that end particle are offering the closest thing to a solution at the moment, that I know of.

I can think of a few approches to your task, but a daresay you will have considred most of them yourself.

I just rummaged in my draw and found the business card for the company that was working in the whitespace spectrum. The website is pretty thin on details, and I have no idea if Paul is still with them. They were aquired by Huawei last year, so they might have cleared out the UK development team and could be sitting on the interlectual property. But I might try and give him a call tomorrow and ask how things have progressed.

Good luck toogooda.

Liam


#15

Mesh Networking isn’t new technology by any means.
There are existing options for Large Mesh Networks.
It’s also possible to use those with Particle’s Backend.
You can select any RF system you want, and push data to a Particle Gateway (Photon/Electron/Boron).
This isn’t plug-and-play, but it’s a viable solution for Commercial/Industrial projects that generate revenue.

However, the main differences with Particle’s Gen3 are (In my opinion):

  • Much lower price point to allow Hobbiest and DIY’ers entry into Mesh Networking
  • The ability to re-program Mesh Nodes (endpoints) OTA
  • The Particle Community (Forum) encourages Collaboration to help advance the System, unlike the existing Mesh solutions in place now. This will eventually make a huge impact in the Mesh Market.

I have Commercial/Industrial Mesh Networks that have been generating revenue for a couple of years.
I also have a box of (25) Gen3 Particle devices sitting in my office.
I intend on helping w/ Gen3 where I can.
I don’t consider either being competition for the other. They are both “Tools” for the toolbox.

At some point in time, this may change. Gen3 has Thousands of “free” developers on this Forum pushing the envelope. No custom solution can compete with that.


#16

From my testing the only real professional solution for enterprise/industrial MESH networking right now are the Digi XBee MESH radio module line. If you need long range then go with the 900HP model of radios to get 1 mile range in urban areas and even more in open areas.

@Rftop is using hundreds of the XBee 900HP radios with the Particle Electons and has been successfuly for years.

I have 3 XBee 900 radios running now with Particle Photon and P1’s for some testing and they are publishing messages and message receipt confirmations continuously back and fourth at a rate of 4 per second with zero issues which is nice.

If you do not need long range and similar range to the Xenon radios you can try the newer XBee3 radios with Bluetooth since they should provide the same functionality just at shorter ranges and the cost for those is around $15 each. Just use them in API 2 mode.

I’ve played with the LoRa RFM95 radios in the past and they do work and are cheaper but are not as well supported or not as easily to get what you want from my experience compared to the XBee modules.

Either way you go you will be supporting Particle considering they have the best platform for IOT that I have seen and these radios still need at least one Particle device to get your data pushed out to the web.

Once the Partice MESH is running stable it will be a very good and short range mesh link system with the easiest setup experience of all the options.

Either way you go some Particle devices should be in the mix.


#17

Yes I do , I have looked at this post and I think the sooner people are directed to it the better so we can see things are boing done. Our industry is legendary for superhero level commitment to solve somthing but getting no credit due to poor communication. The posts look like they are days old becuase the old posts are getting edited.

That’s a good note of feedback. I’ll be sure to relay this in addition to doing daily updates.

This is actually exactly what I am looking for “We are investigatiing to see if it is …but can’t garentee … will sort it for you” It makes me think how can I help. As long as you are clear like you have been in this post it is testing opertunities not false possitives.

I’ll be sure to add there’s interest in communicating more in-between steps in the daily updates.

What is it? New Android App? I am happy to beta test if you need it.

rc26 is device firmware for all mesh devices. This is released separate to the Android apps.

rc26 should help with issues of stability in mesh networks (reconnecting after dropping, UDP packet ordering, etc). We aren’t expecting that rc26 will have any impact on fixing the Android setup flow (although, it isn’t to say that it MIGHT fix something in the flow by proxy).

Android app releases are independent to the firmware releases. I’m not sure of our current status in bluetooth research but I’ll try to dig into that for you this week.


#18

@RWB Thanks for the lead, will do some research.


#19

Hey folks – wanted to follow up here to let everyone know that v080-rc.26 has been released which includes a handful of improvements that address setup reliability, wireless range between devices, and Mesh stability.

Instructions for upgrading are available here – would love to hear your feedback on the update and whether or not it resolves your connectivity issues: