The Internet of Things (IoT) refers to the process of connecting everyday physical objects to the internet—from common household objects like lightbulbs; to healthcare assets like medical devices; to wearables, smart devices, and even smart cities.
The IoT devices placed within those physical objects primarily fall into 1 of 2 categories: they are either a switch (that sends a command to a thing), or a sensor (that collects data and sends it elsewhere).
IoT refers to any system of physical devices that receive and transfer data over wireless networks with limited human intervention. This is made possible by integrating compute devices in all kinds of objects.
For instance, a smart thermostat (smart usually means IoT) can receive location data from your smart car while you are commuting. These connected devices can adjust your home’s temperature before you arrive. This is achieved without your intervention, and produces a more desirable result than if you manually adjusted the thermostat.
A typical IoT system—like the smart home described above—works by continuously sending, receiving, and analyzing data in a feedback loop. Depending on the kind of IoT technology, analysis can be conducted either by humans or artificial intelligence and machine learning (AI/ML) in near real-time or over a longer period.
Think of the smart home example. In order to predict the optimal time to control the thermostat before arriving home, your IoT system might connect to the Google Maps API for data about real-time traffic patterns in your area. It could also use the long-term data your connected car collects to inform commuting habits. Beyond that, IoT data collected from every smart thermostat customer can be analyzed by utility companies as part of large-scale optimization efforts.
IoT often gets attention from consumers, whose experiences with technologies like wearable smartwatches are tempered by the inherent privacy and security concerns that come with constant connectivity. This consumer perspective is prevelant throughout all kinds of enterprise IoT projects—especially when the end user is the general public.
Enterprise IoT solutions allow companies to improve existing business models and build new connections with customers and partners—but not without challenges. The volume of data produced by a system of smart devices can become overwhelming (often described as big data). Integrating big data into existing systems and setting up data analytics to act on it can get complicated.
IoT security is a major consideration when building IoT systems. Still, for many companies, IoT has been worth the effort, and successful enterprise IoT use cases can be found in nearly every industry.
Industrial IoT (IIoT)
Imagine the lifecycle of heavy machinery used on a construction site. Different professionals may stress equipment differently over time, and break downs for any number of reasons are an expected part of operations. Specialized sensors can be added to parts of the machinery that are most prone to breaking or overuse. These sensors can be used for predictive maintenance, to improve human proficiency (an example of real-time data collection and analysis), and inform the engineers who designed the machinery on how to improve new models (an example of longer-term data analysis). Industrial IoT (IIoT) encompasses use cases like these across manufacturing, energy, and other industrial practices.
Logistics and transportation IoT
One of the first implementations of IoT in the logistics and transportation industry involved labelling shipping containers with radio-frequency identification (RFID) devices. These labels store data that can be captured by radio waves—allowing logistics companies to track container movements at certain RFID-enabled checkpoints (like a warehouse or shipping yard). Advancements in IoT have now led to battery-powered tracking devices that continuously transmit data to IoT applications without the need for on-site readers; allowing companies to analyze real-time data for a shipment across the supply chain.
IoT has revolutionized farming in a number of ways, like through the use of moisture sensors. By installing these sensors across fields, farmers are now able to receive more accurate data to schedule irrigation periods. Moisture sensors can also be connected to IoT applications controlling the irrigation machinery itself, automatically triggering irrigation based on sensor data.
Internet of Things blogs
Edge computing brings more compute power to the edges of an IoT-enabled network to reduce the latency of communication between IoT-enabled devices and the central IT networks those devices are connected to.
The ability for devices to compute is becoming increasingly valuable as a means to rapidly analyze data in real-time. Simply sending or receiving data marked the advent of IoT. But sending, receiving, and analyzing data together with IoT applications is the future.
In a cloud computing model, compute resources and services are often centralized at large datacenters. These datacenters are accessed by IoT-enabled devices at the edge of a network. It's a model that reduces some costs and shares resources more efficiently. But, effective IoT requires more compute power closer to where a physical device actually exists.
Edge computing distributes compute resources to that edge, while all other resources are centralized in a cloud. This specific compute placement provides quickly actionable insights using time-sensitive data. Coordinating a fleet of driverless vehicles transporting containers with smart tracking devices is a flashy example, but there are many more practical implementations as well, such as improving healthcare outcomes by analyzing data at the point of care.
Consider RFIDs and the transportation industry: Communication between the RFID and reader is always 1 way. The RFID cannot receive updates just as a central IT network cannot transfer data back to the RFID. It's not a continuously monitoring system, which means logistics tracking is limited to check-ins at certain locations. But if the IoT device could coordinate with IoT sensors installed in the vehicles transporting them, all the data could be managed by the central IT network.
But this connected scenario means each physical IoT device needs a lot of compute power—especially if that logistics company uses complicated machines like driverless cars. Rather than simply sending and receiving—always waiting for instructions from a centralized data center via Wi-Fi—the IoT devices would need to process data themselves and make informed decisions. This implementation of compute power closer to the outer edges of a network, rather than at a centralized data center, is known as edge computing.
As a final example, consider a construction site. Perhaps a construction company has brought a bluetooth-enabled machine to a job site. This machine sends data through workers' smartphones, which helps the company track the machine's use and location. If 10 employees work around that device all day, their smartphones are constantly pinging the cental server—desribing the machine's location. This redundant server activity can overload an IT system. But a mobile IoT app can use the smartphone as a small, low-power server and reduce unnecessary pings back to the central server.