Electronics Weekly Electronics Design & Components Tech News
Figure 1: An example of a smart building system topology (Source: Johnson Controls)
The infrastructure to support the IoT is well established, extending far beyond the unseen legions of servers and data centres to reach right into our homes, offices and factories. Sitting at the edge of the IoT are sensors, gathering data and relaying it back to cloud services, or processing it locally, in an actionable way. These sensors are integral to making buildings smarter, providing a hands-off approach to controlling our environment. The primary drivers for this are convenience and economics; an autonomous building never forgets to turn the lights off when the last person leaves the room.
What started as a relatively simple control system, using occupancy detection and temperature measurement to set heating and lighting levels, for example, has developed in sophistication.
Figure 2: The main factors to consider in a smart building communication infrastructure
The use of artificial intelligence (AI) will ultimately remove the need for a person to program the schedule for a smart building. Simple sensors that detect general occupancy in a large area will be displaced by more sophisticated image sensors that can recognise individuals and provide control in a more personal way.
The anonymous motion detector will make way for imaging systems that can recognise individual faces, gestures and even moods. Audio control, implemented through smart speakers or virtual assistants, is growing rapidly in popularity. As buildings become smarter, their capabilities will expand to provide a more personalised experience that includes access control and other security features. This extends beyond energy optimisation by turning lights off when a room is empty, to include only allowing authorised people into the room, automatically clearing individuals for access to secure networks when inside, and even helping locate items.
Intelligent energy
Today, two aspects of facility management – lighting and heating – represent about 40% of energy consumption. The concept of using occupancy detection and ambient light levels to adjust artificial lighting levels predates the internet.
Figure 3: Mesh networking extends a network and provides routing redundancy
Despite this long heritage, the adoption of connected lighting is in the ascendency and is entirely enabled by technologies that now underpin and advance the IoT.
One key element in this is communication. The advent of wireless mesh networking has made it much simpler and more reliable to connect smart lighting fittings together. Continued advances in power over Ethernet (PoE) technology, coupled with the extreme power savings delivered though LED technology, mean lighting can be powered and connected using a single low voltage Ethernet cable, removing the need for an electrician to install connected lighting.
These connected terminals form an integral part of the smart building network, as each light fitting can act as a beacon for indoor navigation, for example. It also becomes much simpler to include additional functionality to light fixtures, such as occupancy detection, asset tracking, environmental monitoring. All of these features are enabled by multiple sensors that can be integrated into a single, connected device.
Ultimately the biggest gain will be in the energy they can save by operating in a smarter way.
Smart topology
The topology of a smart building system will depend on sensors and actuators, as described and shown in Figure 1.
The microcontroller or digital signal processor (DSP) at the heart of the system will orchestrate numerous sensors and actuators. This will include those used in occupancy detection, environmental monitoring and access control.
Figure 4: A multi-sensor platform, enabled by the RSL10 system in package (RSL10 SIP)
Actuators may include brushed or brushless DC motors to open and close doors and windows, in addition to the electromechanical or solid-state relays used to turn lights on and off. Variable light levels will be achieved using power modulation, such as PWM, and the microcontroller/DSP may be responsible for this.
Connectivity will be a combination of wired and wireless. There is a growing number of protocols that might be used, some that support the same protocols used by the internet and so are accessible directly, and others that will require a gateway. Low power systems, where the microcontroller, sensors and actuators can conceivably all be powered by energy harvested from the environment, have the potential for virtually self‑sustaining control systems.
Important factors to consider when developing the communications network behind a smart building infrastructure include range, power and latency. The weight given to each of these factors will depend on the actual application. Any noticeable latency between someone walking into a darkened room and the lights coming on will be eminently noticeable for occupants, for example.
Local processing will deliver lower latency than relying solely on cloud processing resources to make local decisions. If a sensor is able to decide for itself if someone entering a room warrants the light levels being increased, it will produce a better user experience. Figure 2 illustrates how these factors can influence the choice in wired or wireless technology.
By implementing simple but robust mesh networking (Figure 3), it is possible to build small networks of connected devices that include light fixtures, fans and other assets. This not only provides a way to extend a network far beyond the reach of a single node, it also builds in redundancy, allowing messages to pass through the network using any combination of connected nodes. So if local interference obstructs a message from using one light fitting as a way‑point, the network will automatically reroute it. Most modern wireless protocols now employ mesh networking for this reason.
Multi-sensor platforms
As technology improves, it is becoming more feasible to integrate multiple sensors into a single platform, creating greater value for connected assets, particularly where the main value is defined by its primary function. For example, the primary function of a light fitting is to illuminate, but it also represents a near‑ideal sensor node for capturing a vast array of data.
Figure 5: Energy harvesting techcan be the main source of energy for smart sensors and actuators
By putting multiple sensors into a single asset, you increase that asset’s value immensely. It becomes a critical part of the smart building infrastructure, yet outwardly looks and behaves like a regular light fixture.
The physically small and low power nature of sensors means that a small outline PCB can easily accommodate multiple sensors to monitor occupancy, temperature, humidity, air quality and more.
Using an ultra‑low power communications device like the RSL10 Bluetooth Low Energy radio allows such a multi-sensor platform to run for years from a single coin-cell battery (Figure 4). It is possible to remove the need for a battery altogether and use energy harvested from the environment to power multi‑sensor, connected platforms (Figure 5). This allows smart sensors to be placed almost anywhere in a building. For example, relatively small and unobtrusive solar cells could be used to harvest enough energy from artificial lighting to power a multi‑sensor platform, sending regular data back to a gateway.
Energy efficiency is going to be fundamental to the continued success of smart buildings, both in terms of making buildings more energy efficient to achieve lower power consumption, and in delivering low power solutions with advanced technologies that can make that possible. Energy conservation will be key, from the sensors to the cloud services being accessed. As the number of sensors deployed grows, the granularity with which we can apply control over a building’s utilities also increases, promoting a circle of energy efficiency.
This, however, relies heavily on the continued efficiency improvements in sensor, processor, and connectivity technologies.
As volumes increase, it may even become necessary to employ technologies that have energy independence, using energy harvesting techniques to generate their own supply.
