Internet of Things architects love to talk about edge devices. IoT edge devices are the intelligent sensors and data acquisition devices that collect, filter and aggregate data close to industrial processes and production machines. With increased memory space and computing power, edge devices are capable of monitoring equipment, running data analytics, and moving decision-making authority from the traditional monolithic back-office applications along the edges of a distributed network of manufacturing assets and closer to the point of need, thereby improving responsiveness, process quality, and production yields.
However, real-time process control is not the only consideration in architecting an industrial IoT (IIOT) network. The power of the Internet of Things is in its ability to form a flexible decision-making architecture and to move analytics and decision making between edge devices (for example, for real-time control), and centralized cloud applications such as fleet optimization, as needed.
In my book The Outcome Economy, I proposed a taxonomy of IoT devices, which can serve to determine the level of autonomous decision authority that should be given to edge devices based on their role in the IoT network. The following is an abbreviated version of this taxonomy description.
Activity-Aware Devices
The basic building blocks of the Industrial IoT are single-task devices such as sensors, pumps, valves, and motors. These devices can measure and transmit discrete units of information (a sensor reading) or respond to a simple on/off command (a pump, a valve, or a motor).
The operating model of activity-aware devices is a linear sequence of discrete data collection and processing functions directly related to the task the object is to perform: turn on the pump, measure the coolant temperature, etc. These devices can log commands and data, but they do not possess self-governance capabilities.
Policy-Aware Devices
A policy-aware device is an activity-aware object governed by an embedded policy model. A policy-aware device can sense and interpret events and activities, and respond to them based on predefined built-in operational policies, expressed as a set of rules triggered by events and result in activity streams to create actions.
Many industrial devices, even simple ones, are policy-aware devices. For example, a thermostat has a self-contained autonomous decision-making capability to comply with a temperature control policy. It’s worth observing that an autonomous device does not have to be a complex digital microcontroller. An old thermostat that uses a mercury switch is a good example of a policy aware analog device.
Process-Aware Devices
A process is a collection of related activities that are sequenced in time and space to accomplish a task or a combination of tasks. Process execution rules are used for dynamic recombination of activities to support a broad range of interrelated tasks, sub-tasks, and activities.
The application model of process-aware objects is built around a dynamic context-aware workflow model that defines the timing and ordering of work activities. Work processes (that is, sequence and timing of higher order activities and events) communicate with other devices on the subnet or centrally to accomplish predefined, high-level tasks.
Cold-chain logistics, process automation and control, robots, and manufacturing execution systems (MES) are examples of process-aware applications.
Why Not Edge Devices?
While the value of edge devices is clear, especially when real-time computation and local decision authority are called for, there are other factors to consider when designing an IIoT network.
Clearly, “smart” devices tend to be more expensive than “dumb” ones. Even as the cost of hardware continues to drop as a result of technology innovation and the growing volume of conceded devices, companies in this hyper-competitive space must remain cost conscious. And, in additional to the cost of their hardware, smart connected devices require significantly greater upfront effort in software development and in ongoing device management.
Additionally, data security continues to be one of the thorniest issues impeding large scale adoption of IoT-based solutions. Digital edge devices pose a considerable greater risk of being hacked than their simple analog equivalents.
Finally, restricting data collection and decision making to the edge of the IoT network negates one of the more important aspects of the Internet of Things: the ability to collect information and obtain insight from a corpus of identical devices operating across a broad range of applications, geographies and operating conditions.
IoT Architectural Considerations
When defining the topology of an IIoT network and how much computing and decision authority give to edge devices, consider the following factors:
- Role: How much local, real-time, autonomous governance is required?
- Security: What is the impact of greater security and privacy risk imposed by remote digital devices?
- Cost: What is the incremental cost of hardware, software development and ongoing device management?
- Data Analytics: What is potential value of data garnered from a broad range of edge devices and applications.
Image: Tableau No. IV; Lozenge Composition with Red, Gray, Blue, Yellow, and Black (Piet Mondrian, 1924-1925)