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
Question, I really like the slide controllers and would like to order some, however I am not sure what the benefit is of ordering a 'perk' at Indiegogo as opposed to just wating?
I guess you may eventually get them a bit earlier, but it will still take 7 months and at ~2,5 times the projected price (including shipping)?
Slide API
The Slide App has a ton of functionality in itself, but a lot of you also wanted to integrate Slide with your existing smart home system. That’s why we added an open and documented API as a stretch goal into the campaign. Today, we’re very excited to give you a sneak peek with our first API release. This access is exclusively available to our backers on our website. Visit the Slide API Backer Connect page (http://slide-api.innovationinmotion.nl/api/backer-connect/) and enter your email-address (has to be the same as used on Kickstarter) to access the API.
I'm definitely keeping my eye on this project. I've got two questions though;
- Does someone know if Slide would still work without an internet connection?
- I see that they've send out an update a few days ago via email. Does someone mind re-posting that update here or forwarding a link? Thanks!
Yes, slide will work without internet connectivity
There's a short update on indiegogo: https://www.indiegogo.com/projects/slide-make-your-existing-curtains-smart-gadget#/updates
A more eleborate update with details on the app that's being developed was mailed to backers. It's too long to post, I can forward if interested.
Yeah, I was actually looking for those details when I noticed the update on IG. I'll send you a PM with my email address; would be great if you could forward their email! I'm especially interested in the app and the update on the exterior design!
part one:
They have working prototypes at the office now but nothing I can use yet so I will start on the framework, then fill in the software so it "should" work theoretically and then will see if we can set a date for testing. I am really looking forward to that!
Hi everyone!
Hope you’re all doing well. Kaj and Thijs here with a new update on Slide. We've been making steady progress over the past weeks and we’re very excited to see things coming together. We know many of you can’t wait to get your hands on Slide, so we also want to give a fully transparent and to-the-point roadmap and production planning. Let’s get to it!
Mechanical Development
We’ve worked hard to test and optimize the so-called ‘critical components’ in Slide. As we mentioned in our previous update, sourcing some of these components has taken a lot of our time. Finding the right motor has been of the most difficult components, as we try to find a balance between power, speed and sound. Each new build, with a custom gearbox, takes about 3-4 weeks to be designed, prototyped, shipped to our lab, and tested. The good news is that after multiple iterations we’ve finally ‘locked down’ a motor that we developed together with a firm specialized in DC motor production. That means that in our Bill Of Materials (or BOM), we marked it as ‘suitable for production’. It might not seem like a big deal, but it’s a big step forward for us.
In this photo above you see the layout of Slide’s mechanical component skeleton. It highlights the big redesign from ‘Slide 1.0’ (at the time of crowdfunding) to ‘Slide 2.0’. The entire mechanics are now aligned along a single axis. A servo in the middle (not pictured, but the servo holder is) transmits the movement on the axis (coming from the DC motor, bottom) to either the upper or lower spool. Both spools connect to a ‘spring box’ we engineered in-house. As you can see, the springs are a big and important component of the mechanics. The reason for this is that we want you to keep being able to open and close your curtains by hand. If you open your curtains, however, the Slide Wire would start ‘hanging loose’. These springs prevent that: because the springs are connected to the spools, the wire is always ‘tight’.
When you look at this mechanical breakdown it might look simple, but it’s taken hundreds of hours of engineering and testing to get here. All parts are custom designed and there’s been dozens of iterations in between. We’ve now locked down 90% of this design in our BOM. The 10% that’s still open relates to the connector pieces to hook Slide up to your curtains – more on that below.
GIF above: milling the custom springbox to fit
Electronics Design
Last time we talked about designing our own PCB, or Print Circuit Board. If you remember Slide 1.0, you’ll know how we used different break-out boards connected to each other with jumper wires (e.g. a separate MCU, motor driver, voltage converter, and so on). It made a lot of sense to develop our own board that has all these functions integrated: it saves space, it’s much easier to assemble and improves stability. Since our last update, we have finished designing this board, as well as proto-produced it:
It’s a small board, but as you can see – it carries a ton of features. It also has over 65 (!) different components assembled on to it. The most important of them are highlighted, but that doesn’t mean they’re the only ones that matter. Because we have a 12V motor, 5V servo and 3.3V electronics, a buck converter on the board makes sure we can output all those voltages. That’s easy – but you also need to think about what happens if the 12V motor, for whatever reason, gets a voltage overload. In that case, you have to cut off power the motor, but you want to keep power running to the microcontroller to avoid a reboot of the software. We have designed separated circuits to achieve exactly this. There is also a small secondary PCB that holds the two magnetic hall sensors. It’s mounted next to the spools, which enables us to measure the position and movement of your curtains – also when you move them by hand.
Above: Slide's main PCB connected to the secondary magnetic sensor PCB
Above: Our PCB-designer Marcel going through the schematics and design
Above: Timelapse of PCB proto assembly (a timely process :-))
Software and API
Adding a second microcontroller (or ‘MCU’) was a rather bittersweet decision, because we had hoped the ESP8266 would have sufficed by itself. The reason it didn’t had not so much to do with software, but was rather a lack of input/output ports for the functionality we needed. Because of this, we added an ARM Cortex based MCU to act as ‘second microcontroller’. It basically means we have one processor to handle higher-level firmware actions, such as WiFi communications with the outside world, and another processor for lower-level firmware actions, such as controlling the motor, servo and processing data from the magnetic sensors.
On the software side, this means we had to rewrite part of the code and move it to the second MCU. We had used the Arduino framework to program the ESP, which is not compatible with our secondary ARM MCU. Long story short: our firmware team is busy rewriting the hardware protocols in the C/C++ language, which will be done in the coming weeks.
With our last update we shared the alpha version of the Slide Open API. Tomorrow, we will be updating our API portal (available at https://slide-api.innovationinmotion.nl) with a new release of the API documentation. We're introducing a number of new functions, including functional routine commands as well as creating and changing users (for geofencing) in the Slide memory. Check back tomorrow if you're interested!
Above: Kaj asking our company mascot Soviet for a second opinion on the new schematics ;-)
Above: Kaj and Valentin testing the current sensor for the 'emergency stop' functionality
Business Development
We’ve been working on Slide around the clock since our Kickstarter campaign (and actually long before), but we think it’s super important to take a minute and update you on the business side of things as well. We’re extremely aware there’s so many Kickstarter campaigns that are run as a solo project, often on the side, and many of them go bust. We realize you might be scared this could happen to Slide as well. Heck – one of the campaigns I backed myself in early 2016 has been delayed by more than a year and they just announced – halfway through fulfillment – that they’re out of funds. So it’s not an unfamiliar sentiment to us. In many of those cases (including that particular one) the campaign is a stand-alone project. There’s no retail or go-to-market strategy, there’s no larger business plan.
For us, this Kickstarter campaign is part of the company we run – Innovation in Motion. We’ve been funded before Kickstarter, and we chose crowdfunding because it’s been a great way to see if people are even interested in Slide, and which features are most valued. You guys were the first to get on board on this journey and we’re incredibly grateful for that. As we try to show, the product itself has undergone massive improvements over the past six months (it’s why we call it Slide 2.0) and we can’t wait for all of you to be the first to receive one.
Having said so, fulfillment of this campaign is part of a bigger business strategy for us. Over the past few months, we’ve also been working hard on business development. We’re talking with retail parties as we prepare for Slide’s market launch after shipping to our backers. Our goal is to get Slide to stores in key global markets by Q4 2017/Q1 2018, where it’ll retail for €149 MSRP.
Slide has gotten great interest not just from you backers, but also from potential (big) partner companies around the globe. In order to capture these scaling opportunities we're ramping up to raise a seed funding round next month. This'll enable us to scale into multiple markets and expand the patent on our mechanics, which was formally granted a few months ago, to other regions. We already have several angel investors lined up, but to complete the round we will hold a 'convertible loan crowdfunding' in August. This'll also be an opportunity for you to join us and invest in the company, and we're super excited to tell you more about it in a few weeks time.
Production Roadmap
We know most of you can’t wait to get your hands on Slide. In case you have missed our earlier updates, just a brief recap. After the Kickstarter campaign, we reviewed our prototype with external engineers, and made the decision – early-on – to go for a significant redesign of the mechanics to create a product that’s stronger, more reliable and of higher quality. We could have taken the prototype straight to production, but it wouldn’t have been the best possible version of Slide we wanted to create.
The redesign (especially the internal mechanics) has taken up most of our development time over these past six months. The parts of the Slide main unit are now as good as ready to go. We still have a few concepting steps to go on the connector pieces that connect Slide to the curtain tracks, to make sure they work on as many different systems as possible.
The second cause of delay has been sourcing and iteration on some of the critical components, like the motor. We’ve been working with external parties and their expertise, which unfortunately means that sometimes you have to wait multiple weeks before you’re able to test a new sample. It sucks – a lot – and there’s not much we can do to speed that up.
We now estimate to start shipping Slide units by October, which puts us at about four months delay past initial shipment date. Perhaps we’re just too Dutch, but we have zero intent of telling you ‘next month’ when we know it’s not true. It’s also why we didn’t commit to a new shipment date in earlier updates, because we simply didn’t know. This new date is based on our roadmap, so we have a high degree of confidence:
Updates in the future
We often hear from you that our updates themselves are good, but that we need to give more frequent updates. So far we’ve been updating around once every six weeks. Our approach to updates to-date has been this: to wait until we have enough information that we believe is worth your time, combine it into a big update, and then release.
But we also see that as we move further away from the initial campaign, the need for information and status updates grows. We’ve decided to change our approach, and from now on update you every three weeks. This means that the next update will probably be shorter, and sometimes we might not have as much to share as we’re waiting – for example – on tooling. We hope this way you won’t lose touch with the project, because we want you to be our ambassadors in the future Our next updates will come out on July 23rd, August 13th, September 3rd, September 24th, and so on, rain or shine.
In the meantime, we keep working on Slide around the clock. There’s four of us – Kaj, Thijs, Valentin and Jesper – working (more than) full-time from our lab in Amsterdam, as well as over a dozen engineers in and around the city that help us create a professional consumer-grade product. Or as we like to call is – the best possible version of Slide. We’re actually leaving for Poland at 6AM tomorrow morning, but if you have any questions or if there’s anything else you want us to clarify – shoot us a message and we’re happy to answer.
Cheers from Amsterdam, Thijs & Kaj
P.S.: Last but not least, here's some awesome photos we shot of Slide 2.0
Hello everyone,
Hope all of you are doing well and enjoying your summer. Here in Amsterdam it’s been mainly cloudy and rainy, but that just makes it all that much easier to stay focused on Slide. ;-)
In our last update we said we wanted to increase the update frequency to once every three weeks. That means the updates – including this one – will be shorter than previously, but you’ll get them more often so you’re fully up to speed. Because there’s fewer “groundbreaking announcements” after just a few weeks, we also focus a bit more on the “behind the scenes” of Slide.
What we've been up to since July 2nd
Visiting production partners in Poland
Jesper, Kaj and Thijs spent a week in Poland to meet with different potential production partners in the areas of EMS, tooling and plastic mold injection. EMS stands for Electronics Manufacturing Services and the EMS provider is responsible for sourcing the different electronic components and assembling the PCBs (Print Circuit Boards). The tooling partner is responsible for creating the tools (e.g. the molds) that are then used by the injection molding partner to produce the plastics. Sometimes a company does both tooling and injection molding under one roof. We focused on meeting with a dozen different companies. We had been in touch with them over the past few months, but waited with our site visit until after Slide’s redesign was complete.
The focus of the site visits is to meet the production partner and discuss our requirements for Slide, but also to see the factory and get an impression of how the company works. That last part is actually the most important. Crowdfund campaigns often go directly to Asia and outsource production to a company without having ever visited, only to run into trouble along the way. At the end of our week in Poland we were very happy we did not go that route. Simply put: the quality of the companies we visited varied greatly. Even more: some of the companies that we were most positive about in their online communication were actually disappointing, and vice versa. Some examples of what’s on our internal check list:
• How is the company organized? How many people work there, and how many of them actually work in production (versus sales and marketing)?
• Do they have experience with similar products? Can they show that?
• How do things look on the factory floor? Is it clean? Do they observe proper quality and safety standards, like mandatory ESD jackets for EMS and earplugs for toolers/injection molders?
• Can they actively think along with us in our DFM (design for manufacturing) choices, or are they just a ‘production shop’?
• Are they likely to take this project seriously and commit? Will they go the extra mile?
After the site visits, we kindly thanked some of these companies but decided not to pursue a partnership with them. With the others, we’re currently negotiating collaborations. (Both in Poland and in the Netherlands everyone is on holiday these weeks now so many firms are either closed or working with limited capacity).
.. And even more in the Netherlands
Over the past few months we have also built similar relationships with potential production partners in the Netherlands. We visited them with the redesigned Slide and asked them to give us a quote (price) for the different parts of the production as well. We’re very keen to explore if it is feasible to begin part of the production here (for example, the tooling) and gradually move the production chain to Poland. The closer to home, the easier the communication and the faster we can work, too – but we will have to assess this from a cost perspective as well. We will also have to decide between aluminum and steel tooling.
We focused on meeting with partners in EMS, tooling and injection molding for now because they constitute the biggest chunk of the production with the longest lead times (amount of weeks they need to produce a component). Further on, we will also meet with prospective partners in assembly and shipping.
Optimizing for production
The different tooling partners we met with also provided a lot of feedback on our component designs. Contrary to 3D printing – where you can create any form – plastic injection molding is more complex, as the design needs to be “produce-able”. This involves very small details that you would not notice, but that are crucial to get it right. To illustrate, here’s some of the feedback we received:
Close-up of spring-box. Feedback: “We recommend to increase the curve of the indent by five degrees outwards” (This is a difference of 0.3 millimeters)
Close-up of part of skeleton: “The combination of holes and the long indent limit the possibility to use a sliding die construction. Can we relocate the holing to the outside?”
Jesper and our mechanical engineering partner have been implementing this feedback – and other – into the CAD files for production.
We have made similar optimizations in our PCB (print circuit board), which houses the electronics. This included, for example, deciding on the optimal cable connection between the main board and the sensor board, the type of connector pins for the Slide Dataport, and how to optimize the design for efficient flashing of our firmware at the end of the factory line.
Steps on software: API and ESP operating system, MQTT plans
You may remember from our last update that we added a secondary microcontroller (ARM Cortex M0-based) to handle all hardware controls, whereas the ESP now fully focuses on handling communications. We’ve completed the firmware on the secondary MCU, which had to rewritten in C. We’re still using the Arduino framework for the ESP, but have become increasingly dissatisfied – particularly from a security perspective. Security in IoT devices is crucial and we’ve decided Arduino just does not cut it. We have decided to move our ESP firmware architecture to Mongoose OS, which is also open source but significantly more reliable than Arduino. There is also a native support for integration with Amazon AWS and Google’s cloud services – one of whom we plan to use. We’ve started the porting process last week and this will continue throughout August.
One of the nice bonuses of Mongoose is that it has a very good built-in MQTT broker. This means that we will be adding MQTT support to our API, which is currently HTTP RESTful only. For those of you who have been working with our API – now’s a good time to check it out anyway, as it has just been updated. There is a changelog in the website, especially the routines have been expanded greatly.
Testing, testing, testing.
We spent an enormous amount of time on just testing things. Every single week we run dozens and dozens of small tests on mechanics, electronics and software – and how it all works together. It is probably extremely boring for you to read, but if you want to know how we spend our time, keep in mind that at least half of it is dedicated to building test setups, doing A/B measurements, validating hypotheses and just figuring out what works best. And in practice, we don’t build a new “new release” of Slide each week, but instead conduct countless separate tests. Some sample tests to illustrate:
• The effect of a change in diameter of the wire guidance axis on the wear and tear of the Slide Wire
• The effect of different coatings on the wear and tear of the Slide Wire
• The added value of using metal gears instead of plastic gears in our coupling system
• Limiting vibrations from the motor (and therefore the sound) by using harder or softer plastic grips around the motor.
• Testing a new type of spring that produces the same strength with fewer rotations.
• And so on..
Conclusion
In summary, we have had a couple of productive weeks in spite of the fact that everyone else – including our partners – are on holiday. We are on track with the planning as shared with you in previous update – and as a result, we also don’t have any further news/changes to the schedule, still planning to start shipping by the end of October.
Our next update is coming out on August 13, so stay tuned for that. In the meantime if you’ve got any questions, shoot us a message here or directly and we’ll be happy to answer. Cheers from Amsterdam, Kaj and Thijs
First work in progress
Slide is the first retrofit smart curtain system that works with your existing curtains. It only takes two minutes to set up. Thanks to its wireless connectivity and integration with smart home platforms, the possibilities are endless. And don't worry: with Slide, you can still move your curtains by hand.
(I did read the description for you :-)