The debate: is MQTT holding back IoT?
Olivier Bloch
#IoT Advisor. #IoTShow host. Ex-MSFT. 25+ years experience in building and democratizing complex technologies from Embedded to Edge to Cloud. Open to Board Positions
What if some of the technologies we think are essential to IoT are actually holding us back? Or rather isn't focusing and debating on which technology is better than the other just a waste of time?
This conversation takes a wild turn when two IoT experts I like to call the IoT Watchmen suggest waving goodbye to MQTT and Kubernetes. Intrigued? Curious? Let's dive in!
"The Internet of Things has the potential to save the planet by providing insights that drive action." - Rob Tiffany
Simplifying IoT: A Fresh Perspective
IoT is transforming industries, but are we focusing too much on the wrong aspect of it? MQTT and Kubernetes, often hailed as revolutionary, might be complicating things more than helping in many cases and scenarios. Instead, simpler, more pervasive technologies like HTTP and JSON could be the key to wider IoT adoption.
This shouldn't be just about tech for tech's sake—it should be about making meaningful changes, like reducing water waste and improving energy efficiency. Did you know that up to 45% of water in London is lost due to leaking pipes? Imagine the impact if IoT helped reduce that! And it wouldn't take complex IoT implementations to achieve this!
During this last episode of the #IoTShow, Rob Tiffany?? and Rob van Kranenburg , both long timers of IoT share their perspectives and opinions on this topic. When asked what technology they'd make disappear if they had a magic wand, they said they'd consider getting rid of MQTT and Kubernetes. Not because these are bad technologies, but rather because the ultra focus on trying to make these the de-facto standard for IoT communications and Edge implementations deters from having the right conversations and be open minded to simpler implementations leveraging existing technologies.
Here are some takeaways from the conversation:
领英推荐
Techniques for Streamlining IoT Deployment
Examples of (simple and useful) IoT in Action ??
Could IoT be put to work to move needles?
Imagine a world where IoT not only connects devices but also helps solve global challenges like climate change. It's not just about gadgets—it's about creating a sustainable future. IoT has the power to transform industries, conserve resources, and improve lives, but...
Are we ready to stop wasting time and resources on debating on the wrong topic and are we ready to embrace the change?
What other technologies do you think are ready for a rethink in the IoT space?
#IoTInnovation #FutureOfTech #Sustainability
SmartConnect IoT
2 个月Thank you Rob-Square and Oliver for your candid take … “….up to 45% of water in London…” and many other cities “…is lost due to leaking pipes…” The business model to address these issues is paramount. DM me to share - PS: no new levies and taxes are going to work. Pervasive tech such as HTTP, and the EOIP mantra prevails. Imagine when we moved from restricted/restrictive mainframe days/technology to client server zone. Many protocols and buses came into play, eg, parallel port printing protocol. For adoption, standardization and ease of usage prevailed through to delivering on e-government solutions, to the Internet as we know it. SmartConnect IoT #SmartBoitier standardized edge (last mile) data capture, and streaming. Perversive technology on play.
Software Architect
2 个月I don’t understand how you propose replacing MQTT/JSON with HTTP/JSON, as HTTP is not a protocol that natively supports the Publish-Subscribe model. In this case, you would be forced to use WebSockets, which significantly increases the complexity of the architecture for a microservices-based application, especially when using Docker containers. Additionally, I believe you are overlooking the fact that in IT/OT or IT/OT/IIoT applications, communications often involve numerous notifications (Publish-Subscribe). In this context, an architecture based on an MQTT broker seems to me to be a much better solution for a real-time event bus. It is more suited to the specific requirements of IT/OT and IoT systems, particularly in Edge scenarios where lightweight communication, reliability, and efficient message management are critical.
Empowering engineers to analyze, understand, and optimize by simplifying access to machine data.
2 个月Keep implementations as simple as possible - yes, please. “HTTP/JSON versus MQTT/K8S” - not sure whether this is a battle being fought in the real world or if the two pairs are just examples of more and less established technologies with distinct use cases.
Defining & realizing the technical infrastructure and standards for digital ecosystems and dataspaces | CTO | Gaia-X AISBL
2 个月I would segment the discussion completely differently into two complementary approaches: - technology push - technology pull What we see — and the (alleged) battle of HTTP/JSON versus MQTT/K8S is just one part of it — is the discussion of technology pushing towards adoption. Apparently, for #IoT, this push (alone) has been slowing down during the last years as all the obvious use cases consuming easy-to-find applications of the tech have already been successfully exploited. Regardless of many technologies, actually, so the discussion is kind of mute as core IoT use cases as device connectivity and device management (and all concomitant low hangig fruits like condition monitoring have been implemented (or can‘t b/c of different economic factors). What we now need in IoT is — and this is also addressed in the discussion but mistakenly framed from the tech layer — a creative discussion on non-trivial use cases. And only if they have beend identified including their NFRs, any deeper tech discussion makes sense (aka technology pull). Refarding the tech discussion itself: This looks like a categorical mistake for me: K8S/MQTT is no alternative for JSON, XML or ASCII would. MQTT runs on thin edge devices perfectly fine.
Senior Engineer at eInfochips (An Arrow Company)
2 个月I agree