In Part 1 of this blog series, I spoke about the IoT taxonomy. However, let me take a double click on the “Internet of Things” from a connectivity and technology standpoint. Today, there are really 2 types of “things” in the internet of things:
- An “old” thing (non-internet connected) – Basically, most devices in the world today
- A “new” thing (internet connected) – Some new devices that have native Internet connectivity
What is important to understand is that while IoT is great, there is a cost associated with fitting a device or machine to be internet connected, for instance, in mass manufacturing. While the cost of embedding internet connectivity into devices has decreased dramatically in recent years, every fraction of a cent matters in some types of mass produced devices. Therefore, today, I find that device manufacturers are pondering the value of internet connectivity and focusing on enabling this for higher value devices (coffee machines) vs mass produced lower value devices (a power socket).
To achieve an IoT scenario, it is important that the device has some type of API to work with (whether that is accessible to the internet or not). Most devices have a specification and a set of controls and some are closed (proprietary), while others are open (published). This has been true for decades. When a device isn’t connected to the internet, it can still have the ability to communicate or connect with a local controller (like a Raspberry PI). Lots of old devices may be using Windows CE or an embedded control system that can be connected to another local device with internet connectivity. These “bridges” (typically referred to as “gateways” in IoT speak) provide a cheap way to on-board older devices into an IoT scenario for business value.
Also, what I see emerging is MESH networking. Essentially, the concept of ubiquitous coverage via a combination of macro and micro network coverage points combined with wireless devices acting as relay points. This MESH network falls into the “IoT Gateway” space as a different way of establishing gateways with more modern devices. In the M2M space this is a future growth area since it reduces one of the recurring cost elements of capturing remote sensor date. There are a few companies exploring this space but it is only going to get stronger as Bluetooth enabled smartphones proliferate and Telcos (and others) continue their migration from macro to micro based data networks.
That said, there is one “Device” that is becoming prevalent—the smart-phone. This device is taking on increased capability and may start acting as the primary “bridge” over time. As the dependence on machines to solve business challenges becomes more mainstream, networks will need strong fault tolerance and connectivity in remote fields will require more capability to deal with “Online/Offline” scenarios due to dropped signals or cars driving through tunnels. This requires that software and business processes understand and take “disconnected” scenarios into account as “always on” isn’t an option today and likely won’t be an option 100% of the time, even as networks become more reliable.
Finally, security is an important consideration as you probably don’t want all devices on the public internet. Private networks or closed networks may be required to prevent malicious activity. For example, NASA rockets going to space will probably require a high degree of precision and would likely use a closed / secure communication channel with all their devices.
So, whether you’re using a modern internet connected device, or an older device with a bridge to the internet, as the “Internet of Things” becomes more mainstream, it will become cheaper to derive business value out of IoT.
For more information on SAP & IoT:
Strategy paper: http://go.sap.com/docs/download/2014/11/14b47e23-0a7c-0010-82c7-eda71af511fa.pdf
Please click here for part 3 of this blog series.
VN:F [1.9.22_1171]A Simple Perspective on IoT Connectivity-- Part 2,