Networks, as we know them, have changed over the past few years and there is no slowing down. When I got into the Networking field in 2007, any network was fairly the same: access, distribution, and Core layers. At the access layer, you typically have either computers (desktops) or laptops connected to switches and maybe a few IP phones. Then came the advent of Bring Your Own Device (BYOD) where employees brought their personal devices like tablets and smartphones into the network and organizations had to scramble to ensure that they could provide the same level of service (and security) for these personal devices. And now again, we are seeing another change – the era of the Internet of Things (IoT).
What is The Internet of Things (IoT)
So what really is the Internet of Things? While there is no “standard” definition for IoT, we can try to describe it. Simply put, IoT is the connection of objects (things) to the Internet and these objects can collect and exchange data over the Internet and/or with each other.
Note: It is not clear if IoT devices must be connected to the (public) Internet to qualify as IoT devices or if it is sufficient to be connected to a private network and be able to communicate with other IoT devices. Personally, I think to be able to communicate over a network whether Internet or not is sufficient to qualify as an IoT device.
Let us dissect that definition we have just given of IoT. Firstly, IoT is about things/objects. What kind of things are we talking about? Basically, anything! Anything that can be connected to the Internet can be an IoT device, from wearable devices to smart thermostats to connected vehicles and even smart grids. In another section, we will look at real examples of IoT devices.
Secondly, we said these objects can collect data. To be able to do this, these objects need to be fitted with sensors and usually, the measurements from these sensors can be monitored and/or controlled. For example, a smart thermostat should be able to read the temperature, display it to a user over a dashboard (e.g. smartphone) and allow the user to control the temperature of the room or house. Usually, along with sensors, these objects may also have actuators to carry out a particular function.
Finally, IoT devices should be able to exchange data over the network and/or with each other. This means that these devices must be fitted with a medium to connect to a network. For example, a Nest Learning Thermostat can connect wirelessly using Wi-Fi and Bluetooth Low Energy (BLE).
Real world examples of IoT
Generally speaking, we can divide the application of IoT devices into three categories: consumers, enterprises, and government. Let us look at examples of IoT devices in these three categories.
This is the one we are most familiar with. Examples of IoT devices in the consumer space include:
- Wearables: Smartwatches (e.g. Apple Watch), Fitness Trackers (e.g. Fitbit), Smart clothing (e.g. Owlet Smart Sock 2)
- Smart home appliances: Smart thermostats (e.g. Nest Learning Thermostat), Smart speakers (e.g. Amazon Echo), Connected Lighting (e.g. Philips Hue)
- Connected cars: Over-the-air updates (e.g. Tesla), Driverless cars (e.g. Olli powered by IBM Watson)
Some consider IoT to be even more beneficial to the Enterprise than to consumers. Assuming that the ‘Enterprise’ can figure out what to do with the wealth of data from IoT devices, the benefits are numerous including improved efficiency (e.g. a machine automatically ordering new supplies before the current stash runs out) and reduced operational cost (e.g. adjusting energy consumption based on varying usage levels).
Examples of Enterprise IoT devices include:
- Manufacturing: Asset tracking (e.g. LotTrack), Supply Chain Management (e.g. CargoSense)
- Healthcare: Improved efficiency (e.g. Awarepoint Location as a Service), Digital Medicine (e.g. Proteus Discover)
- Agriculture: Environmental conditions monitoring (e.g. Edyn Garden Sensor), Livestock health monitoring (e.g. ANEMON System)
IoT is also finding application in various sectors of the government. A few examples include:
- Smart Cities: Smart lighting (e.g. Los Angeles Smart Street Lights), Smart Parking (e.g. PARKISSEO)
- Environmental Monitoring e.g. Texas LCRA Flood Sensors
- Public safety e.g. New Orleans
Considerations for IoT
In this section, we will be touching on a couple of things that need to be considered when discussing IoT including connectivity, data storage and analysis, and security.
When we say connectivity, we mean how IoT devices connect to a network to be able to share data. Common IoT connectivity technologies include familiar technologies like Wi-Fi, Bluetooth (and Bluetooth Low Energy), Near-Field Communication (NFC) and even Cellular (e.g. LTE). Other connectivity technologies that are more focused on the IoT domain include ZigBee, Thread, Z-Wave, and Sigfox.
Before we leave the topic of connectivity, let’s think for a moment about network addressing. Irrespective of the protocol being used to connect, IoT devices connected to a network must be reachable via an IP address. This is one of the reasons there is a push for the adoption of IPv6 because IPv4 (public) addresses have not only been exhausted, but the whole IPv4 space cannot contain the sheer number of IoT devices that is projected to be connected by 2020 (over 20 billion according to this estimate by Gartner).
Data storage and analysis
IoT presents a storage problem: where are we going to store all the data collected by billions of IoT devices? Also, depending on the type of analysis you want to do on the data, you may have to decide between cloud or local storage. As this blog entry argues, cloud storage may not be well suited for real-time data analysis.
Closely related to data storage is what to do with the data when you receive it i.e. analysis. There is hardly any point collecting data without being able to translate it to actionable intelligence (except for record keeping sake).
There are a couple of platforms out there that can make IoT deployment easier. These platforms provide several services including application management, data storage, exchange, and analysis. Examples include Amazon Web Services IoT Platform, Microsoft Azure IoT Suite, IBM Watson IoT Platform, ThingWorx IoT Platform.
Security is a hot topic when discussing IoT and this is not without reason. For example, in October 2016, Dyn a Domain Name System (DNS) provider suffered a Distributed Denial of Service (DDoS) that affected major sites like Amazon, Twitter, and Reddit. Upon analysis, it was discovered that the major source of the attack was from Mirai-based botnets. Mirai is a malware that infects primarily Linux-based IoT devices (e.g. IP cameras, DVRs), turning them into “zombies” that can be used to carry out attacks on other devices.
While it may be more difficult to secure IoT devices because of their size and the fact that they may not be installed in a traditional network environment, it is very important that Security is considered during the design of these devices and not as an afterthought. Wouldn’t you want the transmission in your connected car to be cut while driving would you?
Other things to consider when discussing IoT including privacy (what are you doing with the data you are collecting, some of which are personal), software/hardware management (can you update the software remotely or you need to replace the entire device), interoperability (communication between IoT devices), and so on.
Internet of Things offers some pretty interesting applications in making our lives easier like in Healthcare, Transportation, and Agriculture. However, various factors like security, privacy and data storage also need to be considered.
It is also worth noting that things have been connected to networks for ages without the guise of “Internet of Things”.
In the end, before thinking of joining the IoT bandwagon, we should ask ourselves, “Does this thing need to be connected?” If yes, “How connected?”