Watch the video here: https://www.youtube.com/watch?v=HyyIPKJx3Uk
Bettina’s keynote speech:
Hans Vestburg, Ericsson’s former CEO, stated in 2010 that the world would have 50 billion connected devices by 2020. A year later, Cisco echoed the same number and suddenly the figure 50 billion became ingrained in the industry’s psyche – 2020 would see us connect 50 billion IoT devices come rain or shine.
Only more recently have analysts re-evaluated that number and realised that they were a little bit optimistic. In 2017, Gartner analysts, predicted a more moderate 20.4 billion by 2020. (If you look at the table below from IoT Analytics, we seem to be drawing closer to 20 billion, however only 8 billion are from IoT. So, 20.4 billion is still a little optimistic.
IoT Analytics also predicts that the total number of active IoT devices worldwide will go from 9.9 billion in 2020 to 21.5 billion in 2025. But, what exactly an ‘IoT’ device is can be a bit of a grey area. For example, occasionally we’ve encountered Bluetooth headphones listed as IoT devices which is a bit puzzling, but each to their own segmentation (which is exactly what seems to be happening.)
Nevertheless, out of these 21.5 billion devices in 2025, 2 billion are set to be connected though LPWANs in 2025. That’s a huge number of devices connecting via LoRa, Sigfox and cellular IoT (CAT-M1 and NB-IoT).
And we have some good news for LPWANs because during 2019, the market for them saw an overall increase for each of the following networks; LoRa, Sigfox, LTE CAT-M1 and NB-IoT. Cellular is predicted to grow year on year, with NB-IoT forecast to have a CAGR of 49% from 2017 to 2024. This graph reflects the growing demand across all LPWANS, and so the ‘race’ is on.
Three of the key findings from the IoT Analytics LPWAN market report from 2020-2025:
- The global LPWAN market is growing by over 100%
- Asia-Pacific is the largest adopter of LPWAN technologies
- Utilities remain the major market segment of the LPWAN market
If you’re wondering how we break down LPWANs, it’s simply a 92% mix of the well-known networks we talk about in IoT today: LoRa, Sigfox and Cellular Cat-M1 / NB1 and some other lesser used networks.
So, there is a huge demand for the more well-known LPWANs with most of the demand coming from Asia-Pac.
To give you a quick overview of the present state of LPWANs:
- Sigfox is now present in 70 countries with over 700 partners and
- LoRa is available in 143 countries (which is 43 more than in 2018) with 133 LoRaWAN network operators available globally.
- Cellular IoT is not to be left behind though as there are currently 127 commercial networks globally. This is divided between NB-IoT and LTE-M, with NB-IoT at almost 100 and the rest made up by LTE-M.
That’s a huge step up from where we were say, 5 years ago. In fact, it’s a 53% jump in the number of networks counted in December 2018.
When we set Pycom up, we were always asked “which network will win the race”.
Despite the increasing numbers of network providers, it is clear to us at Pycom that no single network can cover everything. The power lies in having multiple networks, where one network can provide coverage if the others cannot.
This continuous hike in coverage means that there is growing confidence in the LPWAN market, resulting in more and more large-scale deployments and larger IoT estates. There have been several large-scale rollouts from different LPWAN providers, such as Sigfox’s 850k strong rollout for NiciGas Gas Meters in Japan as well as their 250K deployment for DHL cages in Germany. LoRa’s also involved in large-scale rollouts: 1 million vehicle trackers for MaxTrack in Brazil and 3.4 million meters across France with Birdz. Cellular IoT is now gaining more and more ground, with NB-IoT being used to track 1 million electric bikes in China and over 2 million smart electricity meters in Sweden are being equipped with both NB-IoT and LTE-M radios.
Here at Pycom, we believe in network co-existence at a device level.
That means that we believe in giving developers the chance to choose the network that they want to work with and see which one works best for their use-case. We don’t say, “you can only use this network!” That’s a waste of breath and no one likes being dictated to. Pycom is all about giving you as many options as possible.
The LPWAN market is predicted to be worth US$211.5 billion by 2028. we get a lot of questions regarding the profitability of hardware with the predominant question being, how can you make money from hardware?
This usual gets followed up with statements like, the manufacturing and production costs create a tiny margin for profit, hardware is hard and SaaS is where the money is...the list is lengthy.
Yet, hardware can be good business and it’s clear there is some market belief that this is true. I should note that even though Pycom started life as a hardware business, we are evolving to comprise of other IoT stack elements. Pycom’s business model is no longer just hardware.
Back to hardware and profitability. You would normally come at IoT with a stepwise progress around the following elements:
- Presenting data in a cloud platform. The platform visualises and presents vertical specific content to a specific target market segment so they can make better decisions and, in some cases, automate altogether.
- The second step in that process is determining which piece of hardware can collect and deliver the data in a current and cost optimised manner at the edge of the network.
So, this makes hardware your second consideration when putting together a full IoT stack and it’s probably still the most complex and time-consuming activity within a project.
According to Analysys Mason, hardware will make up 26% of the LPWAN market value in 2028. So, whilst applications do dominate the market value, hardware is the second biggest money spinner. To summarise, hardware IS profitable.
But what about the LPWAN race I spoke of at the beginning? Well, if we’re really looking to realise all those billions of connected devices there’s only one way to achieve that – accelerate IoT programs because today, IoT pilots still take far too long, in our opinion.
We can only get to all those billions of devices, gazillions in revenue and immense value creation that everyone is talking about by speeding up the way we go about making data collecting devices that communicate back to the cloud.
Back when I started my career at Motorola creating two-way radios, it would take 3-4 years to create each handset and get it to market. It was a mind-boggling experience for me; a fresh graduate full of idealistic enthusiasm working with a load of very clever people. All of us were tied to a super heavy process and the idea that ‘it must be invented here for us to use it’.
Luckily for me, someone in the organisation saw my drive and passion for getting stuff done and put me on a leadership program. This allowed me to move through lots of different business units, and I finally settled in the mobile devices group. It still took years to get one single handset product to market and no-one could successfully deliver the project with an acceptable level of overhead. Needless to say, you should have seen the business heads roll every year for not delivering on KPIs.
Fast forward a few years to my role within a start-up, where we spent 18 months getting a Bluetooth consumer device created and to market. I left before the division responsible for those products was closed down and the business refocused on core services. A wise choice!
I then joined a MVNO as part of the leadership team, where I again witnessed how difficult it is for developers – some of whom are now our customers – to get their stuff to market. We had one ghastly example of a well-known coffee brand trying to get their machines automated and connected. Each machine had a cobbled-together spaghetti junction at the back to make it perform very simple functions such as monitoring and data sending.
Another example was a well-known consumer products group who were connecting up their lawnmowers. They had predicted 1-3 firmware updates per year and as Murphy’s Law would have it, they ended up with 30-something updates in the first few months alone. None of this came as a surprise though, as IoT projects have multiple pain points, from the difficulty of accessing parts and support, networks that are ultimately redundant to the struggle with a scalable PoC. And all of this topped off with being locked into a cloud platform! It’s no wonder that by the end of the project, the only thing driving it is sheer willpower.
(For higher resolution images, please head to the full deck from Slide 2 – https://www.slideshare.net/secret/7HwErEubU73gaR)
Tech was hard and some projects are indeed complex, but we can do something about that. We could see the potential LoRa presented and we wanted to bring that into a multi-network, fast development environment. That’s how the idea of Pycom formed. I say we, because, of course, I didn’t do all of this on my own.
(For higher resolution images, please head to the full deck from Slide 2- https://www.slideshare.net/secret/7HwErEubU73gaR)
Together with Fred, who is your typical dynamo and visionary CEO, we devised a plan to speed things up and remove barriers by creating a disruptive technology platform aimed at developers in small businesses. In our vision, the platform would help developers all over the world to get their projects from idea to market easily and quickly – the aim is within 3-4 months.
They wouldn’t have to resort to covert operations to get hold of the technology they needed to get the job done, regardless of project and team size. We want to remove the clunky, painful, 24-month process of getting to market before they run out of steam and money – which we believe is every inventor’s basic human right.
So, here Pycom is – 4 and a bit years in.
We started with the hard bit of the stack; the labour intensive and lumpy (from a business perspective) hardware. However, we’re progressing. 2020 is the year we’re releasing a brand-new cloud device management platform that integrates the whole IoT stack including LoRa server, cellular airtime, open source hardware – the works.
We invite you all to join us. Tell us how we can improve your development journey…let’s see each other at the launch after-party after 6 months, not 24!
For the full slide deck, please head here and begin from Slide 2: https://www.slideshare.net/secret/7HwErEubU73gaR