This is the forum archive of Homey. For more information about Homey, visit the Official Homey website.
The Homey Community has been moved to https://community.athom.com.
This forum is now read-only for archive purposes.
The Homey Community has been moved to https://community.athom.com.
This forum is now read-only for archive purposes.
Comments
It’s August 13 which means this is Thijs & Kaj with a brief update from Team Slide here in Amsterdam
For those of you who've been following our updates, you know we've committed to shorter updates – but every three weeks – so you’re always up to speed on our progress with Slide. We’ve got some ups and downs to share on our hardware engineering, software dev and the mobile apps. Let’s get to it!
Summer holidays are bad for business
Summer’s in full swing - though obviously, we're taking our break a bit later in the year to keep the speed up. Unfortunately that doesn’t hold for most of the partner companies we work with. In our last update we talked about our visits to production partners in Poland and the Netherlands. We hoped to have selected a partner for plastic mold injection and PCB assembly by now, but prospective partners in both countries are either working on low capacity or even shut down for multiple weeks. Because of that, we haven’t been able to finalize a partnership yet. (There’s still plenty of communication to and fro, but nothing final yet). Based on our experiences, we are more inclined to work with partners here in the Netherlands, but it’s also important to find a partner that can keep the speed up. Hopefully, we’ll have more by the next update.
The good & the bad on the engineering side
Just like last update, we have been spending most of these past few weeks testing several subsystems. Because we already talked about it at length I won’t repeat all that, but a few results are worth sharing.
On the positive side, we’re very happy with the new springs we tested (and talked about last time) – they’re very powerful while small in size. As a result, we’re also able to place an extra spring in the Slide connector piece (that connects the Slide Wire to the inner end(s) of your curtain) to make sure the Slide Wire is always tight and never hangs.
We also found a new combination of a wire – that can hold up to 40kg (or about a 110lbs) – together with something called a linear bearing that massively reduces friction. In our wear & tear durability tests we were able to increase the expected life span of the wire from three years (with the previous wire, and before adding the linear bearings) to about 20 years. It’ll probably last even longer, but that’s as far as we tested.
We do want to be honest and stress that it is not all sunshine, though. We’re super unhappy with the continued tests of the servo – the small motor that operates the coupling mechanism of the open-loop spools (that hold the wire). It is too unreliable, and we decided we need to find a better solution. We’re now going to test an alternative where we replace the servo with as small DC-motor.
IoT security comes at a price
Another issue was the implementation of security protocols on the ESP8266, the microcontroller we use in Slide. In our previous updates we already mentioned that we added a secondary microcontroller (ARM Cortex M0 based) to handle all the peripheral controls. We also said we were migrating our firmware architecture from Arduino to the more professional Mongoose OS in order to meet our security standards. This migration is ongoing, and it will take until the end of September to complete, in time for the beta testing of the device.
However, as we worked through this it turned out the ESP8266’s computing power for encryption is sub-par. Handling a so-called ‘TLS-handshake’, required to set up a secured connection to our server*, takes almost 20 seconds! That’s disappointing, and frankly, not good enough. Because of that, we’re upgrading from the 8266 to its more powerful big brother, the ESP32. This means we’re adjusting our PCB (print circuit board) design to prototype a board version that has this new chip onboard. The main advantage is that the ESP32 has hardware acceleration for encryption on-board, which brings down the connection time to less than a second. Moreover, the cost difference for this upgrade is acceptable. As we’ve said before, security is key for any IoT device these days. (On a cool side-note, the ESP32 also has Bluetooth on-board which means we can add BLE connectivity in the future through a firmware update – but it’s not something we focus on for now).
*Server: In our last update, some people got confused when we mentioned the cloud functionality. We want to reiterate that using our ‘cloud connectivity’ is optional. It allows us to offer features like geofencing and control of your Slide when you’re not connected to your WiFi network. It is not required to use this functionality if you do not want to, and you can just control Slidedirectly within your local network.
The Slide App got a facelift
In parallel to the hardware engineering, our app development team has been working hard on building the iOS and Android apps for Slide that will be available in the Play & App Stores when Slide releases. A few months back we already shared some rudimentary screens that showed the overall user flow of the app. Most of the development on the apps happens in the back-end (e.g. handling the API), but since the last update some of the screens have been improved to add a sexy look & feel to the functionality. Below are some screen captures the team wanted to share with us. It’s not final yet, but we feel it already looks a lot better! What do you think?
Closing remarks
As you can read, we’ve had some ups and downs in the past few weeks. In spite of the summer hiatus, we’re still on track to start sending out our initial beta batch by October. The lead times on some of the electronics is quite long, but among other things we’ve placed component orders for the initial batches to make sure this doesn’t affect us and to align with our planning.
We hope this update is useful for you, but please let us know what you think. We always try to be transparent and increased our update frequency because of that, but keep in mind three weeks is not that long in the world of hardware, which is why we keep the updates short. We’ll be back with the next one on September 3rd, but if you have any questions in the meantime, shoot!
Enjoy your Sunday, cheers from Amsterdam,
Kaj, Thijs & the rest of team Slide
With all solar panels attached, I can't look outside anymore. Obsoleting the blinds and curtains
Hi everyone, Kaj and Thijs here with a Slide update from.. Berlin!
(Or actually, writing this from the back of a driving car on the road from Berlin to Amsterdam
). We were in Berlin these past few days to visit the IFA, Europe’s biggest consumer electronics fair. We attended IFA to learn about the latest smart home developments as well meet with prospective partners from around the world. As always, here's our three-weekly update with the latest on Slide.
IFA Berlin: The Smart Home Era is coming
At Innovation in Motion we already follow everything around smart living, but even we were blown away by the amount of IoT and Smart Home products and solutions demonstrated at the IFA. Some key take-aways:
Our overall observation was that there were too many smart home platforms and connectivity solutions. We saw over twenty different platforms that claimed they will be ‘the center of your smart home’. We really believe that standardization is needed, because the smart home space is highly confusing for the average customer now.
In addition to seeing all the new trends and developments, we also met with prospective retail and distribution partners from around the world – from North-America to Japan. As we prepare our retail launch for Q1 2018, it was interesting to learn how smart home is being adopted differently in different parts of the world, with varying business models and cultural distinctions. All in all, it was an insightful and productive visit.
Hardware process
If you read the last update you might remember we were struggling with the servo that drives the coupling mechanism. Good news – over the past few weeks we have tested stronger servos that do work. This means we won’t have to do further experiments with alternative solutions (e.g. with the small DC motor), which saves quite some time.
As part of the finalization of the internals, a large number of smaller optimizations has also been made: a more robust power connector and reset button; larger magnets for more reliable position measurement, and so on. We have also attracted an extra mechanical engineer to join the team, who is focusing on the mounting mechanism. As the internals are being locked down for the initial production round, the engineering focus shifts to optimization of these external components.
Slide API/connection security enhancements
As mentioned we are undergoing a major architecture upgrade to the ESP32 microcontroller with the Mongoose operating system instead of Arduino. In doing so we focus on the security of Slide’s connectivity on the local network and to the internet. For those of you who have been working with the API, you know that for the local network our REST-like API was HTTP based without authentication. We’re now announcing that the API will be enhanced with an authentication protocol called Digest Auth.
We have been planning to add some form of authentication, and considered using basic auth (which simply means adding a user and password parameter to each JSON call). The problem with basic auth, however, is that if the connection is compromised just once, all communication is exposed. Digest Auth on the other hand establishes a shared secret key between Slide and the user, and uses a hash authentication with each command. So even if the communication is intercepted, it cannot be abused by someone with bad intentions (e.g. a hacker). Moreover, no password is stored on Slide. We’ll provide more information on digest authentication with our next API update in a few weeks.
Further Slide App UX improvements
In our last update we showed some screens of the ongoing Slide App redesign. Since then, our designers have continued their efforts to make the app better looking and easier to use. Here are some screen captures of the new designs focusing on setting routines:
In the screen on the left, you can pick a routine (combining text and icons). In the right screen, you see the sunrise routine which allows you to set an offset. Here, the user set the curtains to open an hour after sunrise.
These screens are for setting the alarm routine. The background color changes based on the time of day. As you Slide through the hours, the colors adjust automatically.
Here’s part of the ‘timer routine’ flow. As you can see, you can set the curtains to take up any position – not just open or close – using the slider. The curtains are, of course, animated.
Wrap-up
That’s all for today! And, because we know you want to know – yes, we’re still on planning for our first round of beta units to be shipped at the end of next month.
This round will be very important to get first feedback from the field as well, and make sure we’re on the right track with the product. More on that in next update. In that update we will also include a video demo of the unit in use with the app – we actually wanted to do that this time, but didn’t take IFA visit into account.
As always – feel free to reach out with questions and comments here or at thijs[at]goslide.io. Next update: September 24.
Cheers from the A1 highway close to the Dutch/German highway, Thijs, Kaj and the rest of Team Slide.
Seriously, these guys are an example of what a better Kickstarter project looks like; clear and recurring communication, transparant in timelines, they listen to community feedback and a bit geeky.
The devil's in the details, of course. So we still have to wait if they can deliver on the expectations - but it's looking good so far.
Hi everyone,
Hope you’re all doing well. Kaj and Thijs here from a surprisingly sunny Amsterdam with a new update on Slide! The two of us actually just returned from London, where Thijs was invited as a speaker at the Smart Home Summit to discuss consumer pains in smart home products. The Summit is one of the key industry events of the year, and it was a great opportunity for us to meet with different market leaders and discuss potential partnerships, including Google and Nest. A great opportunity to put Slide out there as we continue to build momentum.
For this update we decided to do things a bit differently: we got a lot of requests to show more instead of giving just a text update, so today we’re putting out a video update. It’s a nice opportunity to show a lot of things in action and hopefully get you excited too.
So: enjoy!
Video: https://vimeo.com/235246766/2cbaed755b
Behind the scenes look on the upcoming first (beta) batch
If you’ve been curious about where the work on Slide happens, this video will give you a nice sneak peek behind the scenes. Be warned though: our office is (always) a mess.
We’ve started preparing for the first batch of beta units we will be putting out among our backers next month. We will put 50 units in the field to collect feedback from ‘real world’ users and ensure the product works as expected. For this round, we are still using our (high quality Zortrax) 3D Printers to produce the plastic components of the Slide. In the video, you’ll see what this looks like and the manual labor involved in preparing these parts for assembly.
Our initial plan was to start tooling earlier. Tooling is the process where you create molds (in aluminum or steel) that you subsequently use for a process called plastic mold injection. In other words, literally the tools through which you produce your plastic components. In July we visited a number of toolmakers and mold injection companies in Poland, but unfortunately no fruitful partnership has resulted from that. It turned out the companies we visited were either too small for us (e.g. simply incompetent), or so big that they did not take our order volume of 5000 units seriously. Luckily, the contacts in the Netherlands were more willing and we are now working with a toolmaker over here.
Tooling for plastic is very costly, and we were advised by every toolmaker not to start tooling until after ensuring with 100% (99%) certainty that the specifications of the components (shape, size) are right. We simply could not afford a design flaw and start all over with tooling. In other words: everyone told us to first put the initial beta batch in the field, collect feedback, and then freeze components for tooling. So that is what we’re doing, and it’s why our 3D printers are currently running 24/7.
Slide & app demonstration
In the video, Kaj gives a brief demo of the redesigned Slide unit controlled through the Slide App. He also walks us through the user flow of the app, including zones and setting different kinds of routines. The hardware is also demonstrated, including an update of the mounting mechanism that is more robust than the one we used in Slide V1. The mounting mechanism is not yet finalized – we are actually testing different versions in the upcoming beta batch to compare which works best in the field.
Some of you have voiced concerns about the size of Slide V2. Indeed, Slide is quite a bit longer than in the initial design. However, it is also slimmer (less wide), and importantly the thickness is still the same at 5 centimeters. In other words, it’ll fit behind your curtains just like V1. The overall volume increase is only about 20%, and this was necessary to increase the durability of the mechanics. And remember, Slide is always hidden behind your curtains.
One of the things we’re very happy about is that Slide is very quiet when the motor moves. We created a rubber-like mount for the motor that absorbs most vibrations, and the chassis isolates the rest of the sounds. In other words, you will hear the curtain runners make the same sound they always do, and in 9 out of 10 cases that will make more sound than Slide itself. Unfortunately it’s difficult to capture this in a video. At a later stage, we’ll try and do a lab test of the decibel production, but for now you will have to take our word for it.
Next steps
Our focus for October is to ship out this first beta batch of 50 units. As explained in the video, for this test we have two groups in mind: (1) ‘normal’ users that are not particularly technical, to understand how they use and experience Slide and the app and (2) tech-savvy users, including some working with our API, to thoroughly test all functions. From both groups, we are especially interested in how the mechanics work with all kinds of different curtain systems, rails and rods. It will be an important point to determine if it works as expected, or if we need to make changes. For logistical reasons we prefer backers in the Amsterdam/Randstad NL region (so we can come visit and inspect problems in person).
The timeline after that depends largely on how this beta batch is received. If all is good we can proceed with tooling and scale up production fast, if we still need to make improvements based on field results, we will have to do that first. Again, the absolute goal is to get a good product in all of your hands, not a half-working one. We’re all working around the clock here, 7 days a week, but we will focus on a high quality product, not shipping a mediocre product fast. Slide is a new type of product, there’s nothing of its kind yet, so bear with us as we pioneer this field.
So that’s all for now! Do let us know in the comments if you think the video was of added value, so we know if we should do this more often. And stay tuned for our next update, which will come out in three weeks on October 15.
Cheers from Amsterdam,
Thijs, Kaj and the rest of Team Slide
@jorden gezien onderstaande update van Slide heb jij nog beta toegang gekregen?
Hello everyone,
Hope you are doing well. Kaj and Thijs here with a new update on Slide. Apologies it’s a few hours into Monday – we spend all of yesterday going through the beta round requests (so many, great!), hence the delay. It’s also a bit of a shorter update as we faced a big setback with the new circuit board (PCB) which took up a lot of our time, but as we said before: rain or shine, we’ll keep you in the loop. Let’s go!
Struggles with the new electronics
As you may remember from the previous updates, we decided a while back that our architecture at the time, based on the ESP8266 and a secondary Arm Cortex M0 MCU, was falling short. It could not provide security (e.g. fast and stable HTTPS) that meets the current IoT standards. Because of that, we upgraded to the ESP32, which has a much faster dual-core processor and hardware acceleration for encryption (TLS/AES). This switch required a redesign of the PCB, as the architecture changed a lot with all our hardware now driven through a single processing unit.
We completed this redesign last month and started building prototypes of the board right after, but unfortunately we ran into a lot of unexpected hurdles in the process. Without going too much into detail, the bulk of the issues was related to how the ESP32 drives the hardware (e.g. the servo and motor driver) compared to the previous STM chip. For example, PWM was handled differently and due to current behavior we had to do a lot of trial and error with resistors to get it right (and burning a few of them in the process). All of this was further complicated by the fact that we’re simultaneously transitioning our firmware from Arduino to Mongoose (more below), and sometimes we could not see if a problem was caused by the new firmware or the new hardware.
Tech talk aside, long story short: lots of frustrating setbacks with the new circuit board, which took us weeks and weeks of testing and adjusting to fix. To end with a positive note – we do have the hardware up and running now, and we’ll actually be producing the first batch of 60 of these PCBs in the coming two weeks for the initial (beta) batch. Here’s a photo of the old and the new board, next to each other:
But software is running great
As the saying goes – ‘you win some, you lose some’. As our electrical engineers struggled to get the electronics right, our firmware engineers were able to power through with the new Mongoose OS-based firmware and that’s been going well. The development happened in parallel, using ‘break-out boards’ that simulated the new hardware architecture, which allowed the firmware team to cover a lot of ground before the new board was ready. Mongoose is running stable and we’re very happy – it’s a welcome change from Arduino. We now have HTTPS authentication handled in less than a second, and are rolling it out across both local and cloud APIs.
For those of you working with the API – we’ve been holding back a new release as there have been a lot of changes to the new API structure over the past month. It’s now finally stabilizing – with just a few more calls being defined – and we expect to release an update on the API portal shortly. The biggest change will be that the API is now RPC Based. We’ll have different authentication layers, using Digest Auth for all Slide-to-Server communication but allowing ‘simple’ basic local HTTPS calls if you prefer so. JSON RPC is protocol-agnostic, which is a fancy way of saying it will be easy for us to add MQTT support along the way as well.
Upcoming beta round
We got almost a hundred emails from people after our last update asking to help with the upcoming beta round. Thank you! It’s really great to see such strong interest from people wanting to help with the next step. We followed up with a survey to get some insights into the different ‘curtain situations’ as well as use environments. We’re focusing on people in NL, because we want to be able to conduct site visits if people encounter issues with the hardware. Due to the high number of requests we can’t include everyone in this round - we hope that makes sense to you.
As we showed in our last update we’ve been busy getting all materials for this first series – this is still ongoing, and we will spend a lot of time in the next few weeks building the actual units here in our office. We’re also using the opportunity to test a number of things concurrently – including the mounting mechanism. Some units will be fitted with the mounting mechanism we showed in our last update, but some will have an alternative mount that placed Slide directly underneath the tracks. Aside from countless tests here in the office, we look forward to see how both mounts perform ‘in the field’ before we scale up production.
In closing
So that’s all for now – not our best weeks with the PCB setbacks, but we’re making steady progress. In our last video update we showed Slide in action, but only on our single demo curtain. This led to some questions as people wondered how this works with double curtains. As most of our backers have double curtains, we obviously developed with this in mind and shot a quick video to demo it. (The demo video features a prototype without casing, but you get the idea).
(Don’t mind the mess..
)
We’ll keep you posted on progress over the next few weeks, with the next update coming on November 5. Have a good week everyone!
Cheers from Amsterdam,
Thijs & Kaj
Hello everyone,
Hope you all had a good weekend! Kaj and Thijs here with the latest on Slide.
Today in addition to the regular update we have not one but two videos to share! One of the videos is about PCB production, the other one demonstrates a 6 meters double curtain setup with manual control.
What have we been up to?
First and foremost, we’ve been sourcing, printing and building Slide components for the beta batch. That involves by and large very uninteresting work, including hours of removing support material from the 3D printed components, milling and drilling the aluminum skeleton and frame, and so on. Electronics is, however, worth elaborating on a bit more.
As you probably remember during our last update we talked about some of the setbacks with the electronics, specifically the redesign of the PCB (Print Circuit Board) that did not go as smoothly as hoped for. Luckily, this time it’s a good-news show! With the setbacks out of the way, we’ve sourced the components in bulk and given the go-ahead to our production partner in the east of the Netherlands to produce and assemble the first 50 PCBs for our upcoming beta round. Thijs and Valentin visited on-site during one of the production days last week and shot some video material to show you what ‘EMS production’ looks like. Check it out here:
For those of you that can’t watch the video or otherwise don’t like videos altogether, a quick textual recap as well. Personally, I find the production of electronics, specifically assembly, one of the more magical aspects of building Slide. We’ve visited numerous EMS production sites over the past few months, but you never get quite used to seeing a pick-and-place machine meticulously placing thousands of millimeter-sized objects on a PCB in a matter of minutes.
The process starts with so-called ‘barebone’ PCBs – boards where the circuits are placed, but without any components. In our case, we have two such PCBs for Slide: the main board that contains almost all components, and a small sensor board for the magnetic positioning sensors.
Using a custom stencil, solder paste is placed exactly on those parts of the PCB where components must be placed. The pick-and-place machine than rapidly fetches the dozens of different types of components and placed them on the right spot. (In our case, we had two P&Ps, one for smaller and one for larger components.) Finally, the board goes into the oven which then heats the solder paste over the course of 7 minutes, after which all components are firmly in place. Finally, a small number of components must be soldered on manually (so-called ‘through-hole’ components). Fun fact: each Slide PCB has over 90 different components, many of whom you can barely see!
Next up: beta testing with backers
With the final components arriving in the next few days we can finalize the assembly of the beta batch and start delivering them to the backers participating in the beta. For those of you that are in that group: we’ll email you on Tuesday with a proposed date and time to visit you to deliver Slide and ensure installation proceeds without problems. We expect it’ll take about two weeks to get all beta units up and running in the field, because the availability of most tester-backers is limited to evenings and weekends.
During the beta test we’ll monitor and measure a wide variety of things, including which functions and use cases are most often used, how the app works, how the hardware works, the user experience of the onboarding process, and more. We’ll elaborate a bit more on this in the next update. Now our next step after the beta batch depends entirely on how the beta goes. Because Slide is a new product, we’re very keen on resolving any product flaws before we start mass production. With about 50 units in the homes of backers, we’ll be able to determine if we can scale up to mass production, or if issues arise that need to be addressed first. This is complex hardware, and we do not want to start shipping at mass scale prior to completing this large field test.
Big curtains + manual control demo
Over the past few updates we included a few videos demonstrating the redesigned Slide in action. Because we used our (portable) demo setup which is only about two meters wide, some of you were worried if Slide would work with your longer curtains as well. No worries – Slide works with curtains up to six meters. To demonstrate, we’ve recorded a brief video of our 6 meters long, almost 3 meters high double curtain setup.
This is also a good opportunity to talk a little bit more about how manually controlling your Slide-powered curtains works. With Slide, you can still open your curtains by hand. If you have a double curtain setup, you can designate one of your two curtains – the one opposite of Slide – as ‘fully manually operable’. That means that curtain can be opened or closed without affecting the other one. Of course, you can also manually open the curtain closest to Slide, but that will simultaneously move the other curtain too.
Slide itself cannot perform ‘asymmetric curtain movements’: it opens or closes both curtains at the same time, fully or partially. Rest assured, though: Slide’s curtain automation perfectly works even if you've put your curtains in an asymmetrical position by hand. Slide will simply ‘correct’ the asymmetry and either open or close both curtains, depending on your choice.
In case you’re a little bit lost by all this curtain talk ;-) – we recorded a quick video that demonstrates what all this means in practice. Check it out below:
That’s all for today – we’ll be back to you on November 26 with the next update! As always – if there’s any questions, don’t hesitate to shoot us a message.
Cheers from Amsterdam, Thijs & Kaj
Hi everyone,
As always - I hope everyone is doing well.
Thijs and Kaj here once again with a brief update on our progress with Slide here in Amsterdam.
First off – the small practical point that we’ll be doing our regular updates on Monday instead of Sunday from now on, for a host of obvious, practical reasons. :-)
What’s happened in the past three weeks?
To be honest – we’ve had three tough weeks.
In our last update, we showed how the electronics for the beta round were produced, how our software transition to Mongoose OS was completed, and how we were getting ready for the beta round of 50 units. After that, we started by putting the first five units into the field. And that’s when we ran a ton of issues. Prior to that, we had been testing Slide with a variety of curtain setups at our office, as well as in our own homes, and the units worked fine there. But when we put out that handful of units to some of the more challenging ‘use cases’ selected for the beta test, the units wouldn't cope.
As it turned out, we had designed the Slide algorithms to work fine – but they were tailored to the setups we had. When we used Slide with some of the more exquisite test situations, it would not work. Most of those issues were related to the key functions of calibration and peak voltage management.
Internally we had been focusing on double curtains, and wrote the calibration algorithm in such a way that it would make sure the curtains fully close in the middle. When we went to a field case with only a single curtain, we discovered that that same renewed calibration algorithm would put too much force on Slide upon closing. Because of our double-curtain focus, we forgot to check how the same algorithm would perform on a single curtain setup. The solution here is to have different calibration flows depending on if a single or double curtain is used.
Similarly, we use a peak voltage management algorithm that monitors how much current the Slide motor is using, and it acts as a hardware failsafe. It was designed to make sure that if your curtains hit an obstacle, or if the calibration is off for any reason and Slide keeps going even if your curtains are fully open/close, Slide will shut its motor off before causing damage to either itself or the curtain system.
Because we had been so focused on ensuring Slide can handle very heavy and long curtains, we unintentionally overpowered this current failsafe. We would let it reach current values that were way too high for regular curtains. As a result, the wire would snap, because we would let the (very powerful) motor keep pulling it for too long. Someone described it as “using a Range Rover to open your curtains”. Again – the solution here is to adjust the peak voltage management depending on the user’s curtain setup.
In addition to these two algorithmic issues, we also discovered the analog servo (not an unknown cause of problems, for those of you who regularly read the updates) started malfunctioning after about 1 year of (simulated) use. We've now switched to a digital servo which has so far ran for 4 (simulated) years of use – and hopefully this’ll be the last of the servo struggles. But like it, we had dozens of other, smaller, ongoing issues and iterations.
So, basically, these were three weeks of many ‘one step ahead, two steps back’ moments. And to be honest – that’s been tough and exhausting on all of us, especially because we wanted to get started with the beta round as soon as possible. From your point of view, that puts us at the same place (‘ready to start the beta round’) as in the last update. Behind the scenes, we've obviously gone through a complex process of smaller and bigger improvements and fixed. To put that in numbers: we've had 150 internal releases of the new firmware, of whom more than a 100 in the past three weeks alone. We've also had 29 different beta releases of the iOS & Android apps.
Timeline
So, despite the setbacks that we want to keep you updated on – it’s good to end on the positive note that barring any unforeseen but now quite unlikely new issues, we can move forward with the beta round this week.
I also want to re-post a timeline that I shared in the comments before, because not everyone may have seen it and I think it helps you understand what’s next.
Because we've been off with timing before, we are hesitant in mentioning specific dates in the updates. I've been on the other side as a backer plenty of times to know how you feel. And as mentioned, we're also keen to be clear and transparent to all of you, so I'll give you the best estimate I currently can:
We'll need the next 1 to 2 weeks to get all the beta units set up. We want to take about four (4) weeks for testing (including analysis of the findings). That might seem long, but we anticipate some issues may not reveal themselves immediately (we're dealing with hardware here, in different home environments). Assuming we need to make a few changes (but nothing significant), it'll take four (4) weeks to design the tooling (DFM/DFT), and six (6) weeks for tooling itself ('tools' are the steel molds used for plastic injection, to produce plastic parts). This means that if all goes well we can then assemble and start shipping the final, packaged units.
I know it's not as soon as every single one of you would like to have it (and I am *very* aware we intended to ship as early as this summer). But I hope this breakdown clarifies how each different step in the road ahead takes up its own piece of time, and that adds up. As we move forward, and especially as we're getting feedback from our beta users (!), we can refine this planning.
On a more personal note:
Obviously, since we started this campaign, as well as when we decided on Slide's redesign, and frankly - throughout this process - each step has taken us more time than expected. That sucks. But I'm convinced it's worth it. We didn't ship the version of Slide you backed in this campaign, because we knew we could do better. I often hear (frustrating ;-)) things like "oh, your product can't be that sophisticated" -- but it is. Because we're creating something that does not exist yet, that no one's done before - so we’re pioneering. And along the way, we’ll continue to need your support.
That’s it for now – and stay tuned for our next update on Monday December 18th.
Cheers from Amsterdam, Kaj & Thijs
Kaj and Thijs came over, really nice guys who clearly know what they're doing. Had a nice chat with them while they installed the Slide prototype.
It took them a bit longer because my curtains are less than 5 cm from the wall, and I have them in a "cove" that is not optimal for installation, but still fast enough.
After they were installed, I was surprised by the lack of noise it produces. The movement of the curtains is very smooth, and it works well to open them partially as well. The Slide calibrates itself, so it knows how far it can move the curtains.
So far I love it... next step is to make Homey control it
I also got my prototype Slide installed today, so if you have a Homey app I can or need to test just let me know. They did said something about that the current API on the website is an old version (they thought I was also a programmer of some sorts, which I am not) so perhaps you should check if you have the newest one (if you do not already know about this of course).
@lubbertkramer
I will try to make some videos tomorrow! I really love this piece of hardware, software also seems to be quite alright. No disappointing parts so far.
Hey everyone!
We hope you’re all doing well in this cold December month (we’ve had an unusual amount of snow for Amsterdam-standards already :-)). Kaj and Thijs here with the last Slide update of the year, ready to share some good progress with our beta round. And once again, a nice video to go with it showing Slide in action in some of your houses!
Slide beta test in full swing
Those of you who read our last update know we’ve had to postpone the beta round a few times due to setbacks producing and assembling the hardware to put 50 units in the field. Doing a beta round is very important, and although we did not properly schedule it in our initial planning, it became obvious that we had to do it before the costly tooling process (to produce our custom plastic components), not afterwards. So that is, in a nutshell, what we’ve been up to these past three weeks: printing, assembling and installing beta units.
Many of you – our backers – replied to our call for beta testers a few months ago and from those responses we selected a diverse group of backers to participate in this beta test. Some of the testers are highly technical – even building their own software integrations using the API – whilst others are just everyday users. We’ve also chosen both people that live by themselves as well as couples and families, with and without cats and dogs. And of course we’ve chosen different types of ‘curtain situations’: length, weight, track type, and so on. All this to ensure that we collect as much relevant data as possible from this test.
We’ll collect data in two ways: through questionnaires and in direct contact with the testers, as well as by using aggregated data analysis of usage of the hardware and software. Over the past two week of having units in the field we’ve already had a lot valuable insights. Because the prototype unit is not yet complete (e.g. the installation wizard is not yet finalized), we’ve gone to beta testers ourselves to set the units up. On a few occasions, we’d even be unable to set Slide up right away, for example because of the placement of a curtain track or the type of runner. But that’s exactly with a beta round is for – to find these things out, and fix them before scaling up production.
Universal Slide Mount (rails systems)
One of the nice things we’ve been able to include in the beta test is a new universal mount for rails-based curtains. Contrary to earlier mounting designs, this aluminum mount places Slide directly underneath your tracks instead of next to it. It’s attached to one of your curtain stoppers, as well as a runner, and Slide is then mounted underneath. Compared to our previous design, this mounting method increases Slide’s stability, and means you don’t need 5 centimeters of space between your curtain track and the wall/window – but only half of that. The mount is actually designed in such a way that Slide has a high degree of freedom to pivot to the left or right when the motor is turning.
This free-float design ensures that even when the Slide motor is using a lot of power to move curtains, that power is in no way applied onto the curtain tracks themselves, eliminating the risk of bending or damaging the tracks. In addition to fitting in most tracks, there are also screw holes to attach the mount directly to the ceiling, as well as sideway holes to attach your curtain, ensuring Slide remains out of sight.
Beta field test impressions video
We’ve already gotten a lot of feedback per email (thank you
) that we can work with from people’s first impressions. And as Slide is moving curtains in dozens of houses over the Christmas holidays, we’ll learn more about how the device is being used, about what works, and what doesn’t work (yet). We’re very excited to see the first of you interacting with the Slide prototype. In closing, here’s a quick video impression from setting up Slides at beta testers, and capturing some first responses!
https://vimeo.com/247865834/84497baed5That’s all for now. With Christmas on our doorsteps, we’re all taking a (much-needed
) holiday break next week to celebrate the holidays, and recharge for 2018! It means our next update will be in four, not three weeks, and will be out on January 15. Until then, wishing everyone happy holidays from the whole team here in Amsterdam. 
Cheerful greetings!
Thijs & Kaj
Hello everyone,
Thijs and Kaj here with the latest update on Slide. We hope everyone has had a good winter holiday and is back to work/life feeling rested. Here at IIM we've all enjoyed our break and are back to work feeling re-energized. :-)
Holidays or not, Slide didn’t stop moving – literally, with two dozen beta units installed in backer’s homes across the Netherlands and another dozen in our own homes. In our previous update in December we included a short impression video of the beta test start (if you've missed it, be sure to check it out here). Getting the first units in the field amongst you was a major milestone for us after working on Slide for such a long time. Building the redesigned Slide is quite the journey but this was a big step.
And – hooray – the feedback so far is overwhelmingly positive. In this update we’ll focus on beta testers’ experiences and what’s next.
User experiences from the field
Over the past week we surveyed beta testers on their experiences so far: the good, the bad, the unexpected and what they’re missing. We’ll break down the most important findings below.
How often do people use the Slide App?
Average: 7.25 times per week (once a day)
What's the most used routine?
Automatic opening and closing on sunrise and sunset
What has positively surprised the beta testers?
What did NOT go as expected? Disappointments?
Cable visibility (graph)
Would you recommend Slide? (graph)
How did people around you respond to Slide in action?
Conclusions to date & to-do's based on them
To start off, it’s good to reiterate how happy we are with the initial feedback. Since we started our campaign, there have been a handful of concerns on top of people’s minds: “How much noise does it make?”, “Is it intuitive / easy to use?”, “Isn’t the external wire too visible?”, “Will it really work on my curtains too?”. This beta test is an important step in validating Slide is indeed quiet, easy to use, retrofit and that the wire mechanism is not a nuisance.
Obviously – it’s not all sunshine: we gotten some key insights on both software and hardware components that need to be improved. We want to avoid the rather classical mistake of ‘over-engineering’. Especially in crowdfunding it’s not uncommon for a campaign to put out a survey a year after the estimated date asking whether backers would be interested in new features. That’s why we divided the feedback in two categories: ‘need to haves/must-do fixes’ and ‘nice to haves’. Our focus is on getting a good product out there ASAP, so we’re only focusing on the need-to-haves. We've listed the three most important ones below, as well as what we’re going to do next to address them:
1. Servo ticking
Those of you who've regularly read our updates will know that the servo, which is the heart of our spool-coupling mechanism, has been probably the biggest pain-in-the-ass in building Slide. We've gone through dozens of different brands and types. When we started the beta test, we thought we had resolve the mechanical issues. And although the servo does what it should do, even the slightest misalignment will cause it to make an audible, ticking noise both directly after a curtain movement and at random times. Even though the servo now actually works, this is still an issue most testers identified as unacceptable. After a year of engineering with a servo, this is a clear line and Kaj is now working with our engineers on a servo-less coupling mechanism.
2. Connectivity issues
We’ve received a lot of feedback on lost connectivity, to the point that people needed to unplug and re-power their Slide multiple times per week to keep the app functional. The good news is that it does not seem to be a hardware (wifi signal/range) issue, but a software issue. [Technical] Our findings so far suggest that the connectivity issues are related to the way our firmware handles local IP connections – indicating it’s a DHCP/multicast issue [/]. Valentin is working on this together with our firmware partners to find at least a hotfix that allows us to move forward with the current architecture.
3. Curtain system compatibility
This was the area in which we already expected most work to be done: seeing how Slide would actually operate on the variety of systems we mounted it on. The good news is that out of 40 cases (including 25 among backers), there were only two where we were not able to set up the current design of the mount on the curtains. But in many cases, it took us took much time or required additional equipment. Because Slide is designed as a product you should be able to set up yourself, we’ll have to make sure mounting will be easier on a wider range of systems. We haven’t seen any ‘showstopper’ cases, but it’ll take some effort to ensure the mounting system is more universal than it is now (and, obviously, a very good manual).
Next
Over the next three weeks we’re focusing on the engineering to-do’s that are emerging from this beta test. We want to use this period to effectively estimate how much work is left to mass-production so that in the next update, we can provide an updated timeline. As always, we’ll try to be as transparent and complete in our information flow as we can. To be effective, we’ve split up the beta in two groups: instead of installing 50 units at once, we only set up half, and we’ll set up the improved prototype with the second half of the group. We’ll put out our next update in three weeks on the dot, on Monday February 5th!
As a close-out, here’s a video made by one of our beta-testers, Erik Mulder, who has connected his Slide unit to Amazon Alexa.
(click below to watch)
Cheers from Amsterdam,
Thijs and Kaj
--------------
@jorden what can you tell us of combining slide with Homey and how are your experiences?
My curtains rails are too close to the wall currently; less than 5cm. IIM clearly stated that 5cm is the minimum, but I ordered anyways since I have the ability to move them further from the wall. It's just a lot of work
So once they gave me a last minute call if I was around for them to install a beta Slide, I never thought of this anymore and they had an unpleasant surprise - But still managed to install Slide, it just took a bit longer (sorry guys
This means, Slide is now hanging on 2 of my railsliders - a stopper and a normal slider. This is not the way it is meant to be setup, so in no way respresentative for the way it should be. It caused a problem on my setup, Slide moved the stopper and so the whole setup moved a few cm's from the position it had been before. I had to uninstall almost everything and place it back, and since then it's been a few weeks without any problems. I got to know the system a little better because of this
I set up the slide in the iPhone app, which is close to 100% perfect already, and set up "routines": Open the curtains at sunrise, close them at sunset. I changed that this week to 15 minutes after sunset, which I like better. This has worked 99% of the times, I think I have suffered from the software bug that causes a disconnect 2 times now (in around 1 month). I have no doubts that they will fix this before they send the Slides to backers.
My impressions so far:
- Very good app, one thing is kinda unclear during setup (I have reported that to IIM)
- Installation will probably not be < 5 minutes, but I think within 15 minutes should be possible. I do not have an installation manual yet so I had to figure it all out myself when I had to redo the installation, but it was 90% clear and 10% I guessed correctly. Please mind that I have 2 left hands when it comes to stuff like this
- Slide is a bit bigger than expected, but I think that the beta version I have is different than the final product.
- I can fully hide Slide behind my curtains without a problem. Actually, it's quite hard to show Slide to people because I placed the curtains "over" it. It is 0% visible if you look at my curtains, whether they are opened or closed.
- Moving the curtains by hand is still possible and not a problem at all. I haven't tested yet how long it takes before Slide sees that they are open.
- I was shocked by the lack of noise! I was expecting a lot more noise when opening/closing, but it is a really low level of "smooth" noise. Not annoying at all, I can hear it from meters away if everything in my house is quiet but it often happens that Slide has closed or opened while I was home without me noticing.
- The curtains move really smooth, not with shocks/pulls but just one smooth move. Fully opened or closed they hang beautiful, better than when I do it by hand.
- I also hear the ticking described in the mail. It's only noticable when you sit right next to it, and happens at random times. I didn't think of this until they mentioned it in the mail. Will report that
I am really looking forward to receiving Slides for my other 3 curtains. I open and close 2 of them by hand every day, and now that Slide does the other one I am really getting annoyed by doing the other 2 by hand....
The Homey app is also close to done. I am only struggling to make the cards that Homey produces automaticly (Condition: "Curtains are open/closed", Action: "Open/Close curtain") because nobody seems to be able to tell me which function/capability Homey is using for that... but I'm sure Bas will tell me once I bug him again.
The "mobile" card in Homey works though, using a "dimmer" slide to be able to percentually open/close the curtains using the device or using the flow action card. I need the app before summer's here, because then I usually have my windows open and when the windows are open it is not possible to close the curtains. Going to use a doorsensor for that.