Interview #4: Software and cloud by HEXIO
07 April 2021
Source: Software and cloud by Hexio
Nice day to all listeners of our Internet of Things channel, this is Iva from the IQRF Alliance. I would like to welcome my dear guest, Jiri Hybek, technical director from Hexio.
Jiri, please tell us what's new at Hexio.
What's new with us? I can say two things. The first is that Honza Pokorny and I, as the founders of the company, are moving into its management, ie. I am taking over the position of executive director and Honza will replace me as technical director. And in connection with this year's news, what will probably be the most interesting is that we are working on the next version of our IoT cloud 2.0, which will bring a lot of interesting innovations.
So you offer companies some services, something you can help them with?
Our main product – IoT cloud – is a platform that aims to connect hardware manufacturers and integrators, on the technical side, ie. that its task is to connect devices, ensure connectivity, solve the integration of technologies together and actually solve the application layer. The fundamental novelty is that I, as an end user or integrator, can build with a few clicks the end-user application in the Hexio cloud, which I can give to a customer. We want to connect the community, these two worlds not only at the technological level, but also at the business level, and for that reason we also intend to launch a Marketplace, where we want to connect these two groups.
So, for example, hardware manufacturers who can't handle the top layer can ask you and you can help them with that? Including that integration?
Exactly, it can be said that our cloud works as a tool, and our ultimate goal for manufacturers is to be able to use the tool, and build their own application themselves simply by clicking.
You already have a lot of experience with those applications for the industry and IoT, so you could say what you think is missing in the IoT market at the moment.
In the autumn of 2019, we had our own IoTEra conference, some of you may have been there. There, I mentioned two fundamental things that I think apply so far and one fundamental problem that I perceive in the IoT world in general, and that is the fragmentation of technologies. Every manufacturer has a solution. Let's say IQRF has the advantage of working on standardization, but if we're not just talking about IQRF, it's still true that one vendor - one application, the other vendor - another application, etc. Once it goes via LoRa, the second time via IQRF, the third time via Narrowband, the fourth time via the Internet, etc. And it's actually quite difficult to put all this together. I think this is a huge challenge and a problem that the market currently has. The second problem with this is that things are terribly complicated. When I want to implement a project, I want to install sensors for the customer, and it must be fast, I will prepare them, bring them there, install them and it should be solved.
Because, of course, if I install five sensors at home, it doesn't bother me, but if I install 1000 sensors at the customer's premises in several industrial halls, the difference is whether I'll do it in two days or a week. This will be reflected in the costs that the customer will pay as a result. For him, it is economic or, conversely, not. It used to be important to somehow connect devices, visualize them in a Grafana or other dashboard, that was the ultimate goal. I think nowadays, the ultimate goal has shifted a lot to what those customers want and expect, and that's the end application. As a target customer, I will receive the application in which I have made everything customized, ideally accessible via mobile phone. Of course, this increases costs. If the customer wants it customized, he should expect higher development costs. We think that what is missing on the market is a tool not so much for connecting those devices, although even this is important to be able to do it quickly and efficiently, the tool that I am able to give to the customer, the custom application and I am able to implement it at low cost.
I understand it's probably difficult to do it universally so that you can connect devices from different manufacturers on different technologies, platforms and actually do it so universally and functionally that the customer can switch it on, set it up himself.
Exactly, and in connection with this, if you follow trends in the world, so-called no-code or low-code platforms appear.
What does it mean?
I think a lot of people today who make their own website know that they can do it by clicking. They have some drag-and-drop interfaces where they drag paragraphs, pictures, etc. And these low-code, no-code platforms bring this concept. Not only for creating websites but also for creating those end applications. I think that's a huge advantage. At that IoTEra conference, we announced that we were starting to work on such a project, and I can say that after a year of development when we started working on it - we call it Hexio App Engine internally - which is such a lego application where there are elements, such as graphs, which of course has Grafana and other tools, but we are also able to solve individual navigation, individual outputs, event notifications, which are data customized to the customer's needs. I.e. for example, if we make an application for the food industry, where we solve temperature monitoring, then of course those reports and outputs will be different than when we solve predictive maintenance in the industry. We used this App Engine in Hexio all year and we built x different applications, for different segments of the industry.
Can you tell us more about them? These projects are definitely interesting.
One simple project was monitoring temperatures in the food industry, where goods must be stored in refrigerators or freezers. There is also a need to monitor and document the temperature, where the deviation must not exceed one or two degrees for some time and there are control zones to which it must fall, and of course, the monitoring then requires some intelligence over it. When the alerting isn't just about – it exceeded the threshold, and then it got back, but it's about - now it's over the threshold and now I'm going to run some time slot in which it will be measured, if it's good by then, okay, and if it hasn't got back by that time, I'm letting the user know there's a problem. When it has exceeded another level in the meantime, it is actually a different type of event, etc. Something has to be written down, something doesn't have to be. It's more complicated.
In heavy industry, we had a project, and I guess we also presented it at the IQRF meetup in London, where we worked with predictive maintenance of brake systems. These were surface mines with huge excavators, and when something moves, it needs to be braked. These brake systems require some maintenance and can still break. Different types of faults can occur there.
What my partner and I have implemented for the customer is remote monitoring of these brake systems, which evaluates various operating metrics, ie, for example, the pressure in the hydraulic brake system related to other alerts, when actually using machine learning to create models that are able to detect conditions that should not be there. And the customer, the service company, can then handle this and service the equipment accordingly before any fatal failure occurs, which then causes huge financial losses.
What I originally wanted to say before I talked about these cases is that in the IoT world I see that each project is completely different. It's hard to have a universal solution, a universal IoT platform, where we put everything together. That's why we at Hexio build our new IoT product by combining our IoT platform we had with this App Engine, and we create a tool where the integrator, or even the final customer, has the opportunity to build the solution tailored to the end customer. Ie. we can say that we want to offer companies in the IoT world support for IoT technologies and add to that the lego with which to build that particular use case.
And are you able to help with that?
Exactly. Our intention is that the person can do it himself, and make it so intuitive that the average user can handle it, but we are of course ready to offer our partners full support.
I guess it's about the compromise not to burn all the energy on universality, to think of all possible cases and waste months, years to make it excellent in all respects so that the project manager comes and says I want this, this, this, but the question is whether not to solve specific cases concerning the monitoring somewhere in the premises, production, or solve another type of project that focuses on parking or other detection, for example, and then go to those target customers with this product and try to offer it.
Exactly. I think this is one of the other advantages of our solution, that at the moment such applications are prepared, it serves as a template, the integrator can offer it to end customers, profit from the fact that it is ready, use ctrl + c, ctrl + v, adjust it for the specific customer and have a complete product that I am selling in a while.
Who are your target customers?
We don't go directly to end customers, our platform is of course also available to people who have IoT as a hobby to try it out, but our primary customers are integrators and partners who make final solutions and deliver them to end customers. Our goal is to be a tool for integrators and these groups of people, because we ourselves, as Hexio, do not have the possibility to cover the specifics of all possible cases.
Can you outline where you want to go within the IQRF community?
Within the community, we have a sense of inner debt. If we think about what we planned two years ago and we're not there yet, we want to get there. IQRF is a very interesting technology. Over the last year, we have installed it for customers, and what we would like to offer to partners in the IQRF community within our cloud solution is the full support of IQRF network management.
Because when I mentioned here that it is difficult to connect devices, it is not so complicated in IQRF and the moment I connect it to the cloud and I immediately see what is on that gateway and at the same time I don't have to set up anything because IQRF has its a certain standard, so of course it will make the job easier.
Have you seen how to work with sensors in offline mode in the form of so-called beaming? Have you tried it already?
I don't think so. Honza Pokorný solves this, but I don't think we tried it yet.
It's definitely also an interesting part of technology, so I recommend you take a look at it. There will also be a presentation at the IQRF Meetup.
We will definitely like to listen. We know that in the IQRF community there are a lot of hardware vendors who need support, and that's why we decided to offer IQRF Alliance partners much better conditions for the cloud, whether it's a free trial for their customers, or support from our side so that if manufacturers want to supply their customers with the end application, or, conversely, they want to give this tool to their customers, they can count on us in this.
It would certainly be beneficial to do a short webinar on how your application is used so that manufacturers and integrators can see what tool you have and can include it in their offerings.
This will be released probably in April, in the form of a private beta version, early access for selected partners, and we are open to cooperation with the IQRF Alliance partners. If anyone is interested in participating in our private beta, don't hesitate to contact us, we'll be happy to include you on the list, and when we launch it, we'll let you know and invite you.
All materials supporting this version will be released in March and April.
There are lots of materials and instructions on how to use it, on YouTube. So that the use of that technology and the learning curve would be as fast as possible.
If you could highlight the benefits of your cloud solution over other well-known cloud services, what would it be?
I would add your personal approach to that.
Exactly. For Czech companies, for Czech partners, there is 100% Czech support.
Do you focus also outside the Czech Republic?
That product is global, ie we will sell it globally, all over the world. It is fully online, fully digital, we have this option. We focus mainly on the Czech segment and the segment of our neighbors, ie Germany, Austria, Slovakia. We also had interesting connections and experiences before Covid in London, when we participated in the IQRF meetup. So it's also an interesting market. As I say, the product can be offered to the whole world.
Thanks for the interview, good luck and I am looking forward to further cooperation.