Sense + Water flow + other integrations

Wondering if the people at Sense have been in contact with any of the water flow sensing companies?
Kohler has partnered with Phyn and is set to release their version. Grohe has Sense, Moen has Flow, then there is Flume, Flowie and a few others.
They are all competing in this new market and looking for their share. So I’d imagine some of them would be quite interesting in listening to new ideas and even possibly tweak their own products data resolution and exporting.

Sense itself has competition from others, as well as smart plugs, smart breakers and an ever advancing marketplace. My washing machine can tell me how much electricity its using, my shower can tell me the temperature of the water. None are quite a whole home product (yet).

Imagine if Sense partnered up with one of these water flow companies to integrate. Imagine if Sense could definitively say that motor and heat device is your washing machine, your ice marker, your water heater, dishwasher, etc… And seemingly more useful to eliminate false detections.
Not to mention being able to further diagnose any failures. Your dishwasher is using more electricity but less water per cycle, that’s why they are still dirty. The valve in your humidifier failed, that’s why your dehumidifier wont shut off. etc…
I’m sure it would take quite a bit of work to tweak Sense Algorithms to include new datasets, especially from outside sources. But I really do think this is something that needs to be explored.
The smart plug integration topic has been at the top of everyone’s Wishlist forever and isn’t going away. I imagine the engineers have been hesitant to involve themselves too much with the smart plug issue not only because of the resolution of data reported, but the fact that everything would get screwed up if someone plugged something else into the smart outlet. Hence why you can’t label/attach simple on/off plugs. With the water monitors this issue would be eliminated, thus providing Sense a much more accurate and reliable dataset to begin exploring this avenue. Not to mention the much needed collaboration and funding aspects if it all.

Sense labs with it detection of power dips/spikes and basic power quality issues is really nothing new or difficult to accomplish, large appliance manufactures are starting to include integrated motor stall and overcurrent protection/notification.

Sense has it’s place and is a great “tool”, I’ve become much more knowledgeable about my own home because of it. But the severe lack of integration with Sense is a major downfall especially when everyone else is focusing on just that.
Every time I see a topic about integration I hear why it can’t be done, mostly China smart plugs which is understandable despite other users finding a way to somewhat less accurately integrate them by emulating. But I almost never hear about Sense attempting to integrate regarding any of the things I first mentioned. Sure Sense somewhat works with a couple Kasa plugs and partnered up with Schneider to supply their version of Sense but that’s about it.

I just don’t want to see Sense become a thing of the past for refusing to integrate and surpassed by some collage kid on Github or whatever who comes up with a whole home system. Sure any other products would lack Sense’s learning ability, which is great and something nothing else can do, but has it’s flaws. That is why integration is key to making it more accurate and a more of an all around whole home product that would be way more marketable to potential end users in a world where people want immediate results.

3 Likes

I agree and disagree. From where I stand, Sense has done a reasonable to good job integrating with other smart devices. That’s partially a function of my original smart device choices, but also a function of which smart devices are the most straightforward to integrate. Some examples:

  • Sense / Kasa Integration - Once upon a time I bought a couple of Eve Energy power monitoring smartplugs even before Kasa came along. They were nice because they worked with HomeKit, but were painful to pull data from (bluetooth-slow via a special Eve app, and only 5 min resolution). Sense with Kasa was a total breakthrough - easier setup, with good time resolution (every 2 sec instead of every 5 min). The Eve’s still have a place - they can be automated with HomeKit so they fit a couple of nice use cases for me. With the low profile KP115 replacement of the HS110, I can’t see a reason to buy any other power monitoring smartplugs unless you are looking for HomeKit or Google Home integration.
  • Sense / Hue integration - Yes, I went with Hue to solve a couple of strange lighting challenges - mostly the ability to configure several large banks of lights that are on the same wired switch. But it turns out that works quite nicely with Sense for watching your low-usage bulbs, as long as your Sense is on a subnet that has wired ethernet as well.
  • Sense / Ecobee integration - I used to have Nest thermostats, but when Google bought them out, innovation died for a few years and it got much harder to pull out data, so I transitioned to Ecobees. That worked out for me because they remained independent and open, and Sense built the Ecobee Historic integration that reads in my HVAC data to improve HVAC models. I can’t definitely say that that helped Sense do a better job with my AC, but I can say that native detection has improved over the years.
    Sense vs. Ecobee: Round II - #31 by kevin1
  • SenseLink / Sense / Home Assistant - I haven’t used it yet, but it seems like quite a few industrious folks have used this open source solution to pull data from a variety of power monitoring or calculating devices into Sense as emulated Kasa devices. So people with a little code experience can integrate additional devices as long as there is a path to power number.
    Emulated Kasa integration Brings SenseLink to HomeAssistant

I do have a list of additional integrations I wish for:

  • Tesla - Tesla has an informal OAuth 2.0 authenticated API - I’m betting that would give better results for my Model S, than the current detection. My Model 3 is on DCM, so that is accurate.
  • NuHeat - I have 5 floor heating loops each controlled by their own smart thermostats. Once again, the thermostats have an OAuth 2.0 authenticated API that provides power info. Not a huge issue today since I have all 5 on DCM together, but would love to migrate the DCM to other less accessible 240V devices.
  • LG Washer and Dryer - both of these are smart appliances and have an app that gives some measure of power. The LG app is crap, so I suspect that LG can never be counted on for a reasonable API, but I can wish. I have the Washer on an HS110, but would like to move one of my DCMs over to the Dryer one of these days.
  • SubZero refrigerators - My 3 sub-zero units allegedly have a 200$ (each) module that allows them to to be connected to a Crestron automation system for both monitoring and control. The monitoring / control includes power management, including some weird and only narrowly implemented ANSI/CEA2045 standard for smartgrid-ready appliances. The good news is that Sense has detected at least some large percentages of the usage of these.
    Standardized Communications for Demand Response - CTA-2045 Standard

Quite honestly, in my mind, home automation and energy is a currently a mess of conflicting and competing standards. I like the idea of Sense picking off the most valuable and stable ones for users, allowing some roll your own, like via SenseLink, and not trying to support every cheap new smartplug that comes on the market.

I started out getting “smart” with a couple generic energy monitoring plugs for my aquariums, mainly for the notification feature of a power outage, the alert notifications are delayed about 15min but it was much better than nothing. My UPS’s notifications required a serial or USB connection which was not practical. The energy monitoring has come in quite handy and what eventually let me to Sense. After first purchasing quite a few more for various lights and small appliances. I’m not about to go out an replace them all to integrate with Sence. For my use they would be more of a hassle than anything as many are always on with multiple devices behind them Sense has discovered such as some of the heaters. Sense comes in handy for the heaters as I have set an alert if they are on for X amount of time.

The main reason why I brought up the smart plug topic is because of the countless number of threads about them, people clearly want to see integration but understandably it’s not practical for Sense to dedicate any resources to trying to incorporate them into machine learning with so many variables and the poor data they provide.

Hue seems to be in it’s own little world, as they required it’s own hub (bridge) until recently. Now they are compatible with Bluetooth, Home and Alexa.
Most smart lights are compatible with Home/Alexa, Smart Life, IFTT.
Which is sort of where I hoping Sense could find a way to incorporate the integration. No so much into the machine learning but the reporting and in a way is an area Sense could eventually go for a more complete product. It would require some separation from machine learning and Sense to pick up on the on/off status of the devices an IFTT thing of sorts, along with user input of the devices wattage and report/ show up as any other device now, trends, devices, etc… This would be a way for users to account for all the undiscovered and undiscoverable devices. Sense’s existing reporting would be a raw form of sorts, with users being able to select between raw and customized views. If a device is detected through machine learning Sense would default to reporting using that method as Sense is by far more accurate. I know I’m not explaining this the best. Essentially a duplicate platform where users could break down and their Other and Always On. Switch goes on bubble appears switch goes off bubble goes away. Again obviously far less accurate but it would be something. For some users some things might turn out to be surprisingly accurate and if so Sense could take that data to further tweak it’s machine learning.

I have a Honeywell Lyric it uses the Z-wave platform which I know little about. My other half works for a large mechanical, electrical, and plumbing supply company. I wanted a Wi-Fi thermostat it was on sale so she picked it up. She’s over on the plumbing side and in touch with a lot of manufactures directly, hence our new interest in the various water monitoring devices. It’s amazing how smart plumbing has become recently from showers that set the perfect water temperature, activate certain heads and even dispense aroma, to bidets that give you that custom clean based on who’s uses using it. Way to high-end for my lifestyle and a bit creepy but nonetheless the way of the future for those who can afford it. I’m still looking to add some simple wifi hose timers to automate my veggie garden watering.

I haven’t either. I’d much rather pay for a complete product as I have no real desire, patients or time to start dabbling into coding. It’s just not my thing. I’m amazed seeing the things people come up with especially on Github which is in part why I think Sense needs to step up their integration game or someone else will, leading to a entirely new product that will suite the needs of Sense’s target end users. People don’t want to use multiple platforms, multiple apps and whatnot to accomplish tasks. They want a simple solution. Which is exactly why Amazons Alexa and Google Home have been so wildly successful as they integrate so many things using into one simple device. Smart Life brings a lot of different devices together under one platform. IFTT is basically what it stand for. Sense could be that niche for the home energy segment if they could just find a way to incorporate others onto the platform. Highly accurate and reliable data that tells you what all those smart home (and non smart) devices are costing you, when something goes wrong and if they partnered up with water monitoring could offer a nearly complete home energy monitoring solution under one platform. As of now Sense seems far too secluded in it’s own little word to make a real dent.

I too had the LG thin Q app for a washer and agree it’s pretty poor. However it does tell you the cycle state, when a load has finished and integrates with IFTT. Energy monitoring is just the number of loads it’s done and it’s filled with useless tips and ads. The basic state data is obtainable, not what Sense needs for machine learning but its there nonetheless.
There certainly is a mess of standards and seemingly non standards, but like you said pick a handful of the most valuable and stable and start there. The most stable are usually the most produced thus the most valuable for the purpose. With that said if Sense could just pick up the on/off state and estimated wattage and incorporate it into a user customized screen until Sense’s machine learning could actually detect it. Users don’t want to wait around for months, years and often times a device breaks down or is replaced before something detected, this would at least give them some sort of value in the meantime and possibly help Sense tweak how it’s machine learning is trained accordingly.

1 Like

I have a Phyn Plus, and the similarity to Sense was obvious.

Phyn is a totally closed platform, no web app (commercial properties have a web management portal), no REST API, nothing, mobile app only. If I’d done better homework I would installed one of the other devices that can integrate with Home Assistant.

I’ve seen recent ads from Phyn and Sense cross selling each other, made me wonder if there is more than just cross selling in the works?

1 Like

Very interesting. With my eyesight a web platform is something I almost need to have.

@obscuredtrip,
Thanks for your thoughts ! I see the promise and shortcomings of Sense as well. But I do see the home automation world as horribly fragmented right now. My home automation is mostly centered around HomeKit / Siri with a little island of Alexa (FaceBook Portal) / Sonos and another island of Ring / Amazon. Was an original Nest thermostat and camera fan and had some Google Homes, but pretty much abandoned because the Google / Nest ecosystem went through several unhelpful upheavals. I have an Android TV but it also does HomeKit / AirPlay. I have played with IFFFT but found it very limited due to the paucity of supported automations, plus the recent limits for the free version. I do hope Sense continues to integrate more sources of power monitoring, as you do.

Consider a heat pump ventless dryer that runs on 120V instead of DCM on LG. Keep the LG for really large items & add something to prevent not just backdrafts but also constant siphoning of conditioned air https://www.amazon.com/HEARTLAND-21000-Energy-Saving-Closure/dp/B00009W3I4
Over time, the idea is to move most laundry to monitored efficient with some random unmonitored vented.
Of course 240V ventless also exist for efficiency + large load capacity.
I’m a little stuck on ease of access to coils for cleaning. There’s a self-cleaning model but I’d still want to clean myself.
The ‘other’ integration I’d like to see is air. It’s expensive to condition & filter and feels great when it is. IAQ or indoor air quality is a topic of interest. Monitors exist but quality, reliability & affordability may not. Someday, a smart home dashboard will. What Is The Living Building Challenge®? - International Living Future Institute

1 Like

@chrisspy, my family has become very familiar AQ particulate monitoring thanks to the wildfire smoke we have had over the last few year. Seems like there are plenty of good networked AQ devices, an we tend to consult PurpleAir, instead of the news, for highly local and real-time AQ (they delineate indoor vs outdoor measurement)

I guess the other thing in vogue due to Coronovirus, is CO2 monitoring, to check for too many people / lack of ventilation. I had a friend give me a home unit, but it’s pretty poor for monitoring - only Bluetooth plus only collects data while the app is going on the cellphone.

I am happy wit our Awair indoor air quality monitor.

This reminds me. I just installed a new Dacor refrigerator at home (Samsung is an owner now, so they’re probably responsible for this functionality…) and there is an API for the version of the fridge with wifi connectivity. The hilarious part to me is that the API is documented IN THE PRINTED USER MANUAL. The calls are described right there after they tell you where to store the butter :crazy_face:

No energy monitoring, though.

2 Likes

We are by no means anti-integration and have developed a variety of things on the back-end to make creating new integrations easier for our team to manage. There are a few things already in development that will release in the coming months, and we have more plans long-term centered around integrations.

1 Like

I haven’t either. I’d much rather pay for a complete product as I have no real desire, patients or time to start dabbling into coding.

Integrating Sense with HomeAssistant doesn’t require any coding these days, you just click the “add integration” button, select Sense, and login. Reporting power usage requires a little editing of a config file, but it’s still pretty straightforward.

I also have some integration with a water monitor, but it’s a little more difficult. Most uses of water don’t have any electrical usage associated with them and flow rates aren’t consistent enough to identify usage.

1 Like

Been thinking about this a bit. Flume in particular. Not just as an integration, but as an acquisition target. The relative company sizes and revenues are right, and it has a good balance of overlapping and non-overlapping technology. With Flume about to take a deep dive into the ML pool, it makes a lot of Sense to me.

I have more “smart” devices that anyone could imagine including smart toilets, BBQ, Fireplace, Driveway Sensors, Pet Feeders, every single light switch in the house, and the list goes on. I also have a Phyn Plus and I wouldn’t integrate Phyn with Sense if I had the option and here’s why. Phyn is TERRIBLE at recognizing what is on. I’ve given up on Phyn completely and only use it for water leak detection and the ability to manually turn off the valve if I get some strange report. I would “train” Phyn to things like my water conditioner, which comes on every day at 2am. It recognizes it for a week then it says it’s a shower. I change it, it stays ok for a while and then it goes to being an “outdoor spigot”. It’s the dumbest smart device I own. If it was connected to Sense it would only rub off it’s stupidy on to Sense.

I run HE and HA as well. If you want to see a combination of the data I would suggest using HA as others mentioned.

1 Like

Okay, I have to draw the line right there…

:slight_smile:

You have no idea what you’re missing. When your toilet recognizes your voice, loads your customizations for music, heat levels and I’ll spare the rest of the details…

Might I add that you can even have it let loose interesting sounds when your guest takes a seat.

2 Likes