Like the majority of the guys and the girls in the office is fresh out of college. Just learning as we go. Three years later, four years, or five years later we rebuilt some kind of expertise regarding software in Automotive, especially with something that we call a remote continuity platform.

I think the main advantage, that literally you don’t have any fear of failure, because you don’t know what’s possible or not. Things just happen. Or if they fail, you try to fail fast, and just keep on doing something, making good things.

So Rimac Automobili had a mission of building the world’s fastest electric car. Somewhere on the road we figured out that we cannot do this in a conventional way. Then we started developing our own things. Other OEMs in the industry figure out that they can use some of our components that we are building. After that we that we figured out that… hey, this is actually our core business.

Should I do a moonwalk or something? So, we have almost assembled Rimac Navera here, and what is important to state regarding the connectivity platform. So, you see here this is our infotainment unit. It’s connected to the 5K networks, and it’s able to transmit up to six and a half thousand signals to our connectivity platform for live data, and for historical data analysis.

Network connectivity is always an issue. I don’t see it is solvable in near future, because you have too many factors in it. It can be problem of the telemetry control unit itself. It can be problem of the antenna availability of the mobile network, that you’re in roaming, that you’re not supporting something, that you have not global coverage. Your connectivity to the server back-end system, data aggregation, databases, are services that are actually doing something useful with the data. The consumers like front-end systems or mobile apps. Everything in that chain can go wrong. We had issues with all of these things. That’s why I kind of know what I’m talking about.

So this Nevera has up to 6,000, 7,000 signals, all of these sensors some of them might be on the tires on, the rearview mirrors, LiDAR sensors. They are getting sampled with sample rates up to 100 times a second, one thousand times a second, depending on whatever you want. Why we want to even get into, let’s say, all of this information from these signals into the cloud… You can actually have, let’s say, infinite amount of processing power, that you can just do anything with the data. You can just like do a lot of applications, like entertainment wise, and let’s say medical wise. You can do safety wise applications without compromising anything or anybody.

After the idea of our CEO to have the vehicle information on the web, we started developing everything from scratch, right. So we developed our own telemetry protocol. It kind of worked. Then, let’s say, extend it a little bit, extended a little bit, and in every software development you know how it goes. You build something on top of it, you build something on top, and then it falls apart. So we had this overhead of maintaining something, that was not our core product.

So we made a really quick and important and valuable decision to switch from our own protocol, and to utilize MQTT.

We actually decided that HiveMQ provides us with the whole base structure of the MQTT protocol without limiting us in any way.

The biggest advantages for HiveMQ, we have a reliable and scalable solution. We tried out some other brokers. But HiveMQ is the only one that we really felt like it’s a seamless integration. We have some features out of the box which are really necessary for us. Like making sure that you can transfer data from HiveMQ to Kafka. Like with just a configuration file that’s like five lines long.

It’s just amazing when the data is sent from the vehicle, that we don’t need to bother that it will not be received, that it will not be aggregated, that we will lose it along the way. And we have support from HiveMQ team whenever we need it. If we wouldn’t have such a reliable partner as HiveMQ, we would drift our focus from the platform itself and its features, which at the end of the day is our core value.

That’s one of my favorite migrations actually, because it worked seamlessly, like until 9 a.m. everything was using the old protocol, and after 9 a.m. everything was using the new protocol, there’s no data loss, nothing. Nobody even understood that we literally implemented a new whole new back end. So one of my proudest periods at that time, yeah.