IoT and AI are two of the hottest topics in technology, which is a good reason why enterprise technologists need to understand them. The two technologies are very symbiotic, so planning how they can support each other is important to help enterprise users.
What is IoT?
IoT is a network of devices, not people. IoT applications are typically built from devices that sense real-world conditions and then trigger actions to respond in some way. Often the answer involves steps that affect the real world. A simple example is a sensor that turns on some lights when activated, but many IoT applications require more complicated rules to associate triggers and actions.
The messages that represent triggers and actions/commands in IoT flow through what is commonly referred to as a control loop. The part of an IoT application that receives the triggers and initiates the actions is the center of this loop and is where the IoT rules reside.
The control loop is only part of the overall information flow in an IoT application – the part that actually receives information about real process conditions and generates real responses. Most IoT applications also generate some business transactions. For example, reading a shipping record at the entrance to a warehouse could open the gate for the driver – a closed-loop decision – and also generate a transaction to stock the goods represented on the record – a business transaction. Decisions made in the control loop must meet the application’s latency requirements, often referred to as the length of the control loop.
Control loops often require only simple processing to close the loop and produce a real-world response to an event. An example of this is entering a code to open a gate. In other cases, the processing required to make the decision is more complicated. If processing needs to apply more decision factors, the time required to make those decisions can affect the length of the control loop and the ability of the IoT to deliver the expected capabilities. A half-minute delay, for example when a worker has to scan a list before admitting a truck into a freight yard, could reduce the capacity of the yard. IoT could read a QR code on the manifest and make the necessary decisions much faster, speeding up the movement of goods.
What is AI?
AI is a class of applications that interpret conditions and make decisions similar to how humans respond to their senses, but without direct human intervention.
There are three general forms of AI in use today, namely the following:
- Simple or rule-based AI is software that has rules or policies that relate trigger events to actions. These rules are programmed, so some people may not recognize this as a form of AI. However, many AI platforms rely on this strategy.
- Machine Learning (ML) is a form of AI in which the application learns behavior rather than having it programmed into it. Learning can take the form of monitoring a live system and linking human responses to events, which are then repeated when the same conditions occur, either by analyzing past behaviors or having the data provided by an expert.
- Inference or Neural Networks Use AI to build an “engine” designed to mimic a simple biological brain and draw inferences that generate responses to triggers based on what the engine “infers” about the conditions. Today, this technology is most commonly used for image analysis and complex analysis.
All three of these forms of AI are designed to represent human intelligence, but their ability to represent something even approximating actual human intelligence is greater if you cycle through the three in the order above.
How can IoT and AI support each other?
In the IoT, real-world events are signaled and processed to generate an appropriate response. Put simply, any IoT application that uses software to generate a response to a trigger event is at least a basic form of AI, and AI is then essential to IoT. The question for IoT adopters and developers is not whether to use AI, but how far AI can go. That depends on the complexity and variability of the real system supporting the IoT.
A simple rules-based AI would say, “When the trigger switch is pressed, turn on light A,” and a more sophisticated evolution might say, “When the trigger switch is pressed, and it’s dark, Turn on light A.” This not only represents the detection of events (trigger switch) but also the detection of states (it’s dark). Programmers use state/event tables to describe how a series of events in multiple states are interpreted, but this only works when there is a finite number of states that can be easily recognized.
Using the example of a truck arriving at a warehouse with goods to be stored, a simple AI could allow the driver to enter a code to pass through a security gate. This would eliminate the cost of hiring a worker at the gate. It is also possible to read a barcode or RFID tag on the vehicle itself and allow access without entering a code. This would allow the truck to continue while its right to enter was validated, further speeding up the process.
If more conditions need to be analyzed to determine a response to an IoT event, the process is beyond the capabilities of a simple AI application. If the it is dark state was replaced by a name, i need more lightand the IoT system should not respond to a specific trigger switch, but to the task that a person was trying to perform, simple AI would not suffice.
In this situation, the ML form of the AI could monitor the arrival of a truckload of goods at the warehouse. Over time it could learn when the drivers and workers need more light and activate the switch without the person having to act. Alternatively, an expert could perform the expected tasks and “teach” the software when more light would be appropriate. AI/ML software would then eliminate the need for a programmer to build each IoT application.
In the inferential form of AI, the IoT application tries to gather as much information as possible and mimic what a person perceives. It then applies inference rules such as People can’t work where light levels are below xand based on the detected conditions and the application of these rules, decides to turn on a light.
Inference-based AI requires more complicated software to collect conditions and define inference rules, but it can respond to a wider range of conditions without being programmed. The same level of inference processing could determine whether additional workers should be used for unloading because the goods are urgently needed, work is behind schedule, or simply because workers are available. All of this could improve the movement of goods and the overall efficiency of truckers and warehouse workers.
IoT is about using computer tools to automate real-world processes, and as with all automation tasks, it is expected to reduce the need for direct human involvement. Although IoT is aimed at reduction in human labor, it does not eliminate the need for human judgments and decisions. This is where AI can step in and significantly improve the IoT system.