In the 1850’s sending high tech messages meant taking your horse and buggy to the nearest Western Union Telegraph station and laying out one dollar for their operator to transmit your five-word message across their telegraph lines. Fast forward to the present day and that pay-per-use, leased infrastructure model has extended through the cellular age of telecommunication carriers.
When wireless technology enabled us to network people via smartphones, and then to network things via IoT, it’s no surprise that the public network carriers were eager to expand their subscription models. They expanded their networks from wired to wireless, paying for bands of licensed RF spectrum instead of miles of cabling, and expanded their subscription models to include data.
Public carriers, however, didn’t win every connection. Wi-Fi emerged as the ubiquitous wireless network for IT connectivity in the home and office. The license-free wireless bands and an enormous ecosystem of standards-based equipment makers have enabled the deployment of private networks, across homes, offices and campuses, allowing users to utilize their own network infrastructure equipment with no monthly subscription fees.
As the Internet of Things, and its enabling wireless technologies, have matured, there is a similar macro-trend developing towards a mix between public and private wireless networking. Private networking, as its name implies, allows for the enterprise to own their network, allowing for a higher degree of control, security and cost containment.
In both cellular and LPWAN technologies, the private networking model is rapidly gaining momentum, while the public carrier model falters. The recent receivership of Sigfox, and the exit from LPWAN by a public LoRaWAN® network carrier in France further reinforce the predicament of these traditional wireless carriers who aspire to own the wireless networks of all IoT traffic, including sensor traffic that generates miniscule levels of data and revenue per connection. That’s a big leap from the traditional smartphone market and raises questions about the downward scalability of the leased network business model to include sensor level traffic.
Why then do private networks have such a high appeal in the IoT?
The economics of “massive IoT” create an ROI problem
It’s challenging for companies to incur additional operational expense every time they add another class or group of assets to their connected portfolio. The return on investment to connect every vehicle in a commercial fleet may be very high, but to extend that connectivity down to every pallet requires a different cost model.
When traditional cellular carriers lacked interest in extending their business model coverage to devices with a budget of a dollar a month, or even a dollar a year, the door opened for new LPWAN technologies to enter the market. Sigfox and LoRaWAN emerged as viable contenders to address the market for low bandwidth, long range wireless sensing networks, but ultimately utilized differing business models.
Sigfox took the public carrier, closed system approach, with pricing that could extend below $1 per month to connect sensors to their network.
Alternatively, LoRaWAN opted for an open, standards-based and adaptable business model. This line of attack supports both private and public network architectures providing the option to choose between a Wi-Fi-like model, where the user buys and maintain their own network, or a public carrier model in which usage or a subscription fee is paid. With Sigfox in receivership and LoRaWAN experiencing rapid growth, it is evident that the market has voted on its preferred model.
Security, Stability and Control
Today’s wireless networks are highly secure but there is a sense of comfort when data never leaves the network. Public networks can suffer from a lack of control, specifically the user’s control. Outside of negotiating and paying for service level contracts with the public network provider, users have no idea if they are sharing leased infrastructure with 2, 20 or 20,000 other devices or how much demand those other devices are putting on the infrastructure. With a well-designed private wireless network infrastructure, users have control over the number, type and behavior of the devices attached.
Private network architectures also offer the ability to leverage different architectures, such as edge-intelligence capabilities where decision making is done at the edge of the network, without requiring that all data be transmitted through the network.
“Things” in the IoT can have very long lifecycles compared to IT equipment and smart phones which rarely see anywhere near a decade of service. The Things connected to the IoT can be deployed for a decade or much longer. The idea of leveraging a public network that may “sunset” during that product lifecycle is a showstopper for many classes of assets and use cases. By choosing a private network architecture, companies take control of their own destiny by owning their own infrastructure instead of leasing an infrastructure where a changing technology or failed business model could leave them stranded.
Choice and flexibility driving success
The flexibility to choose between public, private and hybrid networks appears to be the winning formula for IoT applications. Not every entity can install, manage, and maintain their own infrastructure and will lean towards the leased, OpEx model. Others have the capability to “own” their networks and can take advantage of a more CapEx based model, as well as leveraging the benefits of edge intelligence and a fully controlled network. Some projects that are stuck in the pilot phase have found success by shifting to private network models and a growing number of projects are utilizing a hybrid approach of both public and private network infrastructures. Ultimately, having the option to select an architecture and financial model that most closely fits your business requirements will drive the success and scale of the IoT.