17 September, 2020

10-year-old promise of ubiquitous IoT. Expectations vs reality.

A look at how the core components of the IoT infrastructure evolved and why it’s not as smooth as envisioned.

IoT as an industry has been around for over 10 years now. The Internet of things as a term emerged in 1999 but was popularized by Cisco at the end of the last decade. In 2011 the promise was that by 2020 the number of connected devices would jump from around 13 billion to 50 billion. Radically new levels of operational efficiency and optimization would be achieved in just about every industry driven by grand scale deployments of IoT solutions. Over ten years into IoT existence it’s evident that there are multiple hurdles in achieving the total digitization and ubiquitous connectivity for all things. While IoT still holds an enormous potential, realizing that potential is taking substantially more time than originally anticipated. 

Any IoT solution naturally has three core components — enabling hardware, software platform, and the connectivity that ties those two together. Let’s look at each component individually and review key technology advancement we have witnessed over the past 10 years.

Enabling hardware 

For truly widespread IoT deployments and the financial feasibility of projects, the costs have to be low. Over the last years, we have continually seen rapid commoditization of hardware — everything from sensors to tracking and monitoring devices, which led to consistent year-over-year drops in prices. As global volumes of manufacturing increase fuelled by the demand from a growing number of IoT solution providers, and more suppliers realize higher economies of scale, this will continue putting downward pressure on prices. This will in turn drive more IoT deployments. Module makers —  from Quectel to Telit to Sierra Wireless — have made significant strides in creating modules that combine multiple GNSS sources, cellular connectivity, and even Bluetooth, Wi-Fi, and other technologies to enable device manufacturers create the most versatile devices. Cellular has rapidly evolved from 2G to 3G, 4G and now we witness the first 5G modules and telematics devices. Additionally, electronic components are continuously being optimized for lower power consumption, which increases the scope of applicable use cases. 


The above-mentioned evolution of cellular standards is of course driven by cellular operators continuously deploying new generations of network equipment while shutting down the support for legacy standards. This process is happening at different rates around the world with the US and Australia now almost completely shut down support for 2G and 3G, with the rest of the world to follow. It’s hard to imagine anyone seeing 10 years ago such a rapid evolution from 2G to 3G to 4G and now 5G. With the advent of 4G, IoT-specific technologies that de-prioritized bandwidth and low latency for the sake of cost reductions emerged. Evidence to that is 4G LTE Cat M and NB-IoT technologies. Competing LPWAN technologies in the unlicensed spectrum LoRa and Sigfox further expand the low-cost and low-energy connectivity options. Emerging 5G, on one hand, can offer ultra-low latency and ultra-high bandwidth, but on the other hand, it makes it possible to deploy IoT-specific configuration, which enables ultra-high device connection density per cell at the expense of increasing latency and dropping bandwidth. This will help alleviate network congestion issues as IoT connections grow to a massive scale.

Software platform

Visually appealing dashboards and analytics, access across multiple mobile and desktop platforms, API integrations with various platforms and systems — to name just a few notable trends around software platform components. However, with all these obvious advancements, there’s something that doesn’t meet the eye but does hold the industry back. 

Over many years IoT solution companies have been inventing the same thing over and over again. When it comes to building a software platform for a promising IoT solution the focus is naturally on the differentiated front-end capabilities, such as the above-mentioned dashboards, analytics, user-friendly interface. When we think about back-end — gateways for the incoming data from the connected devices, databases, API mechanisms for data exchange with the front-end and third-party systems — those behind the scenes capabilities get re-created many times and are often not done right. Weak ability to scale into thousands, then tens and hundreds of thousands of devices, deficient APIs that limit third-party integrations as they are becoming more and more important over the years — that’s the unfortunate reality of many back-end implementations, that are built from scratch.

Flespi is the solution that emerged three years ago and is now fixing this problem. Many IoT and telematics solution companies — from micro-mobility start-ups to fleet management providers to predictive maintenance platforms — came to realize that their investment dollars are best spent developing front-end differentiators while back-end capabilities are best outsourced to Gurtam’s flespi platform. It’s evident from flespi’s rapidly expanding customer base that we are playing an increasingly important role in accelerating the global IoT transition.


Looking back at the last 10 years there are certainly many things the industry can be proud of, however, it’s also evident that overly optimistic projections by Cisco visionaries were wrong. By Cisco’s own account last year the number of connected devices we’d have in 2020 would be at 30 billion — 40% below their original estimation. However, what’s also clear is that the growth will not be slowing down. With significant technology improvements in all core IoT components and obvious benefits to business and society, we can expect more investment and more IoT deployments over the coming decade across the globe.