Understanding the smart home ecosystem
This chapter provides an insight in the definition of a smart home, the different values smart homes might add and the different ecosystems that exist. With these insights, the ecosystems are compared and a choice is made for which ecosystem will be the focus of this project. Additionally, a smaller scope of smart products is chosen for more focused prototyping and testing.
Smart home promises
There are many products that fit inside the smart home, but it can be hard to see how they relate to each other and what value they add. This chapter provides an overview and divides the market into the type of value the products promise. This report assumes three main promises: security, energy efficiency and comfort.
Security
The first promise is meant to provide inhabitants with an improved sense of security. This includes smart locks, connected cameras and door/window sensors.
Energy efficiency
The second promise is a reduction in energy usage and as a result a reduction of energy costs. These products mostly control heating systems, some attach to a heater and some replace the wall thermostat.
Comfort
The last promise contains products that make your home more fun, flexible or generally more comfortable be in. This is a broad promise and products in the security and energy efficiency categories might fall within this category as well. This promise contains smart lights, media devices, but also alternative interfaces and sensors.
Closed and open ecosystems
Closed systems
Some of the examples mentioned are products developed by large companies to which services are attached. These companies do not want you to switch services and therefore create closed ecosystems. Communicating between services is only possible when the company specifically builds this functionality.
Millionaire systems
Before all these segmented brands existed, there were fully integrated systems. These systems are incredibly expensive, but they are made to last and keep data inside the home. A company will integrate all your appliances with the system, assuring the devices can communicate.
Open systems
Another way to build a smart home is by using open source systems. These systems are made to be flexible and integrate with as many products as possible. They do not depend on external services or make it easy to switch between services. They focus heavily on privacy and automating behavior, which is likely to become more prevalent for closed systems in the coming years. Figure 1 shows what could be attached to an open ecosystem.
Some large companies do make their products accessible for open source systems. For example, many smart lights use the open source Zigbee protocol. This allows users to control their IKEA TRADFRI lights with the Philips Hue bridge and app or with an open source alternative such as zigbee2mqtt.
Homemade device with a motion sensor and radio control actuator. This device communicates through the open source MQTT messaging protocol. MQTT works over Wi-Fi and allows any number of these sensors to be added to the system.
Problems of closed ecosystems
The service based nature of closed systems raises questions about their longevity, privacy and flexibility.
Service closing down
Homes tend to stay around for quite a while, generally much longer than a mobile phone. Then what happens 10 years after installing smart home products that run on similar technology as a mobile phone? Companies might go bankrupt or decide that providing a service is no longer financially viable. Both of these situations have happened already, with the Pebble smart watch (Verge, 2018) and the Nest Revolv (Verge, 2016) respectively. Because these products have a closed nature, it is generally not possible switch over to another service. These customers were out of luck and their products, while functioning physically, no longer work because the services were shut down.
Privacy concerns
Placing products in your home with microphones and cameras can be detrimental to your privacy. When these products are connected to an external service for processing, this loss of privacy becomes a certainty. A recent study shows that consumers have reluctantly resigned to the idea of giving up privacy for convenience when it comes to smart speakers (Lau, Zimmerman, & Schaub, 2018). This does create tensions between users of smart speakers, which shows not everyone likes to give away their privacy for convenience.
Not flexible
The last issue with these systems actually builds on the previous problems. Once you buy into a system, it can be hard to switch to another system. Many of these products cannot easily communicate between each other. This requires additional hardware or adding another external service. This could force you to depend on even more services that might shut down or share your private data with yet another company.
Problems of open ecosystems
Compared to closed ecosystems, open systems seem very promising, however there is one big issue. The flexibility that open source offers, also makes it incredibly complex. An open source system might have a friendly interface for regular usage, but is generally unfriendly when it comes to configuration. For example, the open source hub Home Assistant uses a programming language to configure devices or to configure automatic behavior.
Complexity can exclude inhabitants from configuration
This creates the situation where very few can configure these systems, generally just one person in a household. This excludes most inhabitants from changing devices or altering automatic behavior that does not for work them. The result of a system that is too complex to alter can be seen in a study conducted in the United Kingdom, as described below.
Study: no energy savings
This study aimed to quantify the energy savings of a smart home, but found that there were many barriers (Hargreaves et al., 2018). At the start of their 9 month trial, the participants did actually have energy savings. But as the test continued, the participants got annoyed by automated behavior. They had no support to fix this and they eventually started shutting down most of the automated behavior. After the 9 month test, there were no energy savings.
Complexity of systems
The test made use of products from 2013, at which point there were no cheap universal systems available. Currently these systems do exist with the services previously mentioned and they are rather easy to use. The open source systems have not caught up, especially not in configuration. See figure 2 and 3 for a difference in possible complexity between a closed and open system.
A closed source system, all configured through one hub and app. Devices are limited by the amount of services that can be connected.
An open source system with two hubs and two configuration environments. The amount of devices and services that can be connected is much larger.
Concluding: open ecosystems and smart lights
Understanding of the different ecosystems allows this project to continue development with one of them. Additionally, a choice can be made which products to include in the development scope.
Open ecosystems
This project will address the difficulty of involvement with an open ecosystem. Closed ecosystems have many problems, but these are rooted in business decisions. Open ecosystems have design issues, but with focused development they could rival the experience of a closed system. The open ecosystems also have a focus on automating behavior, something that is not as extensively possible with closed systems.
Additionally, the closed systems are difficult to prototype and test with, as they do not allow tampering in their system. Open ecosystems are in favor of this behavior and have entire supporting communities.
Smart lights
To allow for tangible testing scenarios, this project is limited to the scope of smart lights. Unlike a thermostat or a security system, smart lighting affects a person immediately when they make use of a room. This allows testing scenarios to be limited to first use scenarios, which requires a low involvement of participants.
Smart lighting products are also quite open. Closed source lighting systems can be controlled through open source platforms very easily, which makes prototype development more accessible. It is rather easy to add a light or controller for prototyping purposes to an existing lighting system.