Why should I use nymea?

The lockdown periods during the Pandemic have given me the opportunity to ‘play’ with quite a few smart home systems and whilst I am no expert and don’t have any affiliation with any vendors; it has raised a lot of questions in my mind.

So having ‘played’ with nymea for a few weeks now - I can’t help thinking - “why use nymea instead of any of the other popular solutions out there?”.

I am not a developer so can’t help nymea by creating plugins, but thought it might help to stimulate some debate to help nymea get the community perspective on their product.

Thanks

2 Likes

Hey 3more,

I guess why you should use nymea instead of any other is probably a question that you can only answer for yourself. For me personally, the answer is clear, but I guess my perspective, being a developer of nymea, doesn’t count here. We believe we’re doing a lot of things right, better than the competition, and we’re also aware that there are fields where we’re lagging behind the competition.

If you’re looking for great and fluid user experience, a good combination of ease of use and flexibility, or if you’re looking for privacy, local-first and security, I believe nymea is a good choice. I could go on with praising things I think are awesome in nymea, but I’m having a hard time praising things I did myself and rather would let you - our community - decide on that. According to your other posts, I would believe there’s at least something that got us your attention :slight_smile:

In any case, if you decide that nymea is the right choice for you, rest assured that we’re happy to see you here and value your feedback.

Well as you point out @mzanetti - the start to answering this question is to understand nymea’s uniqueness.

Simplicity - it is very easy to understand and use. A lot of the learning for these kinds of systems is in the UI, and nymea takes all the pain out of this by providing an ‘as is’ graphical frontend. This just leaves the need to connect the protocol and do some magic. Simplicity is key to nymea and enables rapid deployment.

Small footprint and support for multi-devices. The nymea core is very lightweight and the app runs on most devices making it easy and cheap to deploy. This also means it is easy to use as a Kiosk solution and in remote locations.

Very little community development. This doesn’t sound like a good thing, since most smart home solutions have grown into something great because the developer community have made significant contributions. I believe it is a good thing that nymea is still largely company lead, because this means better control over the quality and security of the code. It just puts a higher burden on the nymea company to accelerate development, so that the community can grow as ‘testers and users’ - validating deployments and providing necessary feedback.

Use of Qt - is both good and bad. Good because it provides a consistent experience and because nymea has support for python, c++ and javascript development - it has the potential to be used in a variety of projects/applications. Bad because it is not the kind of environment that the maker community is used to - although I doubt that the Qt element would put people off.

So it is obvious why nymea as a company deploys the software in embedded solutions and whenever edge middleware is required. But this does detract from development time spent on the smart home solution side of nymea.

At one point I believe that Home Assistant had 1700 community developers contributing. Nymea doesn’t have this and so if it is to compete in the smart home space, then it falls to the company to accelerate development, stretching resources. With no real community development - this is a big ask unless there is a specific resourced plan to make nymea a main player in home automation.

So - nymea has a good number of ‘niche’ uses and also fulfils some of the home automation demand. I guess the next part of a review needs to address how to build on these ‘niche’ attributes, what the overall vision is, and where the community fits into this.

It is OK though if nymea just becomes the ‘embedded smart home solution’ and home users like myself get to use it in our projects; but it would be better if it was to become something more!

If I was asked what would be the overall goal for a smart home solution, the answer would be easy - be better that Home Assistant. Or do what Hubitat, IP Symcon or Homeseer do for free :smile:

This would be a big ask for nymea - so what smaller developments would put nymea into a more mainstream category?

  1. Develop a few more protocols - Zwave is obvious but there are doubts about future use - it still has to be included at this point in time though.
  2. Develop an OpenThread protocol and demonstrate that nymea is forward thinking.
  3. Develop a Lora(Wan) protocol.
  4. Push the boundary and further develop Thread into a meshed failover solution - I believe this would be adopted by smart building projects over and above Zigbee.
  5. As part of the above - provide a way to mesh multiple nymea cores, making it useful for use in large or remote buildings and/or large properties. Similar to the way pilight uses SSDP.
  6. Develop OpenMQTTGateway and Zigbee2MQTT plugins
  7. Develop local customisation and integrate iframe and map widgets to capitalise on the last few points.
  8. Provide integration to other systems that major on protocols e.g. Fhem and IOBroker

Some even smaller changes would really help:

  1. Develop magic with more non coding and scene use, allowing rules/scenes to call scripts without coding. This is useful in OpenHab 3
  2. Allow some basic configurable things to be added by the user.
  3. Allow the user to customise the look of widgets e.g. icons, colours etc.
  4. Write an interface to another open source GUI - e.g. SmartVisu/CometVisu - this is probably a couple of days work and opens up the possibility to have a complete 3rd party front end to nymea. There are already drivers for SmartVisu covering Fhem, OpenHab, SmartHome NG, IOBroker and a generic JSON interface. Or co-operate with newer GUI developments like Jarvis, HABPanel and Lovelace.

Hopefully, the community will provide more, but I think the priority should be protocol development and a SmartVisu interface.

Thanks for the feedback. You’re bringing up some good points there. Especially on the protocol sections, zwave, lorawan and openthread seem good ones. We’ll likely get tho those eventually I suppose. For integrations with openMqttGateway or zigbee2mqtt I’d love to see the community stepping up. I think from a nymea point of view we’d rather work on improving nymea’s own zigbee support.

I know the rule engine could use a little tweaking to become even more powerful, I do have some ideas there. FWIW, you already can run arbitrary scripts by using the commandlauncher plugin.

Also for things like SmartVisu, I doubt we’ll get to this anytime soon (I personally would rather spend the time to improve nymea:app) but I can see how this would be a nice feature for the community. Also here I suppose it shouldn’t be too hard for someone to step up and give it a go? I would certainly help with getting started if someone takes this on.

1 Like

Hi,

Yes I think you are right in differentiating what nymea develops and what the community should develop. I have tried to outline the simple stuff.

For example, given nymea’s interest in the energy market I would have expected Lorawan capability to be a must, but then again you are probably on this in the background. Nymea’s capability as a Lora gateway for instance would be an easy win in that vast business area. There are so many more areas where nymea’s edge capabilities give it an advantage in the business arena, but much less in the smarthome market.

That said, we (the community) get to enjoy the benefits of using it in our homes (static and mobile) on the back of your business - so I won’t complain :slight_smile:

1 Like

I note that the new 4K Apple TV now has Thread built in (Google Nest already has it). That now means that all the major manufacturers are heading towards thread based bridges (border routers). What does this mean - co-existence of current Smart Home hubs and new CHIP approved ones. It also means that most manufacturers will have to establish a thread based bridge and deal with non cloud based solutions - which I think is a big step in the right direction. The new Samsung Smarthings hub now has zigbee and thread built in. Not sure where this will leave Z-Wave but my guess is that Zigbee/Thread, Lora/NB and Wifi 6 will be the dominant protocols.

Why am I so excited? Thread and CHIP certification means more privacy, more resilience and more security than existing Smart Home solutions. Why is this good for nymea? Well it means that nymea could become easily become a CHIP bridge and Smart Home solution in one, possibly beating Home Assistant to the post? :slight_smile:

Thanks for all your sharing. I think we don‘t need a copy of HA, it is already there.
I like to envision nymea as „HomeKit for non-Apple devices“. Focus is good look/feel, reliability, overall quality and privacy obviously. :slight_smile:
All your points are super valid.
I‘d be happy to have a more powerful community. Do you have an idea how to make this happen?
Regarding CHIP / Thread: Sure. will happen. CHIP is a good move but also a bit slow in providing clear specs. :blush:

Agreed, the CHIP process has been slow although I think this is because most of the manufacturers are represented so probably lots of committee work. I don’t think we will see any certified products until later this year of earlier next year. The main point is that these represented manufacturers are already building products with CHIP in mind. I know there are some suspicions around the standard e.g. because it is an application layer built on top of ip v6 - suggests that they expect most devices in the future will be connected to the internet. All speculation and the nymea team are probably much more up to speed on CHIP than I am. My reason for raising it is because I am a big fan of Thread and also because I think it is or will force z-wave to open up their protocol.

As for a more powerful community; they evolve out of need and common purpose and can’t be built. In my book the community follows the product up to a certain point and when there is enough activity/users they then drive the product. I know I keep mentioning Home Assistant, even though I don’t ever want to use it - but it is 4.5 times more popular than the nearest rival and more than 10 times than other open source solutions. So there must be something we can learn from them :slight_smile:

As far as I can tell the Home Automation market is accelerating and the big names Google, Apple, Amazon etc are all pushing easy to use, good looking and competent applications on the back of their hubs and devices. They want it easy for the average person to setup a sophisticated HA solution in minutes. They have the power and money to dominate the market and I don’t like that!

Following in their footsteps are the commercial HA solutions offering a ‘maker community’ like experience with reliability and hybrid company/community support. I don’t know if they will survive the dominance of the big 3. Samsung Smarthings offered this approach and in my view they just ended up with a few spun off solutions. Quite a few others are dropping out.

That just leaves the open source solutions that survive because of all the nerds and makers out there. Some though are now approaching near plug and play status and may cross over into the commercial arena. Some have reached their peak and are now reliable, fun products not really going anywhere. There is also a handful of great vertical solutions that will keep going as long as their champion stays with it.

So, I often find myself ‘wondering’ where nymea fits given that is has elements of all these different types. This must make it highly versatile and although it is just my personal musing’s about what it could and should be. I like nymea, I like it a lot simply because you have achieved your vision, but then again your vision is not much different to that of the big 3 :wink:

With these ploughing 10s of Billions of dollars into their product (after all the most obvious use for voice assistants is to control stuff) and into AI it will be an exciting future. So when my drinking buddy says to me “so why mess around with open source stuff”, my response is obvious; “because I want my system to do what I want it to do and not what a company expects it to do”, long live customisation,! :grimacing:

I like nymea, I like it a lot simply because you have achieved your vision, but then again your vision is not much different to that of the big 3 :wink:

We have by far not achieved our vision. But I think here we have common ground: nymea shall be as simple as the big 3 (minimum) but open source and trustworthy.
There should be a 4th simple and good option, along with the big 3. The world deserves it. And it is just fun to build it. :stuck_out_tongue:

@Simon - it is a good point you make. I guess this is what I mean by being versatile. Just like the big players you aim to provide a relatively simple no-nonsense, reliable, well supported and great looking solution. You even go as far as embedding your solution in other products to provide smart home functionality - just like the big players :slight_smile:

So it goes like this:-

“Hey - we are the makers of ‘she who must not be named’ and because we have the best Voice Assistant in the world we are going to make it the best smart home hub with our version 4 tech. You can connect and control your home by voice, by phone and use our simple, great looking app. Our VA will also learn so that it knows when you are asleep, at home, working or playing with the kids. You can control it from anywhere in the world. And of course…should the internet ever be down we might even let you have some local control from your phone or local stored voice files. And by the way - thanks for all the data - it was a pleasure to really get inside your home and your life routines” :smile:

So along comes nymea:-

“Hi, if you are fed up giving up your life story to the cloud but want a simple, reliable, no-nonsense smart home solution that is fun to use and cool to own, try nymea. We integrate with all the common devices, we are even inside a few, but we always respect your privacy. Our locally controlled Voice Assistant does what you ask it to do and doesn’t spy on you but still provides that great experience of voice control and if the internet goes down - you won’t even notice. We don’t ask for much hardware or additional apps and run anywhere even in your shed, greenhouse, campervan, caravan, outbuildings - all connected, at virtually zero cost. Even when you are on vacation - maybe on a beach or in the van - a secure link back home means you are in full control anywhere. And if you do deploy nymea in your outback vehicle, the great looking display will not only give you full control of the utilities in the vehicle/van but a full picture of what is happening at home. Hey, and our cool local VA goes with you to help make sure your van knows what you want. We don’t need your data, just your loyalty and of course if you want to customise the way it works - the tools will match the time you give it, get under the hood of nymea without having to break open your expensive VA device”.

So as you say Simon - maybe the 4th solution is better because it is ‘the’ private fully functional voice enabled smart home solution.

Just a bit of fun to write this - not as much fun as you are having to build it though! I think your company is going through a very exciting time at the moment :slight_smile:

We need to focus more on the offline voice assistant part I guess. Let us know what kind of books / articles / content you published or are going to publish!

‘In my book’ is a phrase local to me, meaning “in my thoughts/opinion”. I am not writing any books :slight_smile:

You have a choice of open offline solutions:- Mycroft when they eventually integrate deepspeech, Rhasspy/Sepia and Linto. My favourite is Rhasspy especially as the developer is happy for anybody to do anything they want with it, commercial or otherwise.

Dhirn has released a Rhasspy module for Fhem, and I believe there is one for Jeedom from Kiboost? I have had success using a script in nymea but the lack of user defined variables/tags means you have to add every Thing that you want to control into the script.