Smart living, not smart devices, Part II: State of the technology
After Part I that examined the state of the smart home market, this post reviews the way how advanced technologies are being used in today’s smart home solutions.
In many ways, technology is created without the use cases, and users come even later: we have a tendency to overestimate the impact of tech in its early days (along with its concomitant adoption curve) and then underestimate its long term impact on society. This is certainly true for the core technologies surrounding smart home devices. Processing and display technologies, networking, sensors, and artificial intelligence capabilities have experienced revolutionary improvements and cost reductions since the beginning of the smart home journey. However, the current state of smart home devices and services still exposes serious gaps in our ability to put those technologies together into solutions that consumers will embrace broadly. This section explores the gaps in how smart home solutions are composed from current technologies.
Complexity, lack of standards, maturity level
The focus on individual devices and the lack of standards have led to fragmentation that often stymies attempts to connect IoT devices to build useful solutions. Complex installation processes, and diverse and inconsistent user interfaces can make it difficult for consumers to get started with their new devices. Competing wireless network technologies, from Bluetooth to WiFi, Zigbee, and Z-wave make require hubs to allow interconnection. We will explore this area in a future white paper. However, it should not require an engineering degree.
Often, the technical immaturity of features can result in a lack of predictable and reliable function, which will also put off users. For instance, the motion detection focus of a video camera at the front door must be defined precisely to capture potential visitors, but to avoid capturing the motion of the hanging flower pot nearby. Immature sensors and their configuration often result in false alarms or lack of action, making it difficult to rely on them. In some cities around the US, there are even false alarm fees.
The user interfaces of many devices expose individual device features rather than supporting the role they are expected to play in consumers’ lives. “Universal” user interfaces, mainly provided as smart phone apps, often lack expressive power for the broad set of functions they are expected to control. Many user interfaces consist of deep menus of device functions and conditions that require home owners to become programmers, and still don’t permit the full expression of intent in specific situations. Depending on the function, any of a range of other interface modalities may be more suitable to control actions than smart phone touch screens: voice commands, gestures, simple presence, or the output from sensors such as light sensors, thermometers or cameras.
True adaptation to individual situations requires an ability to express users’ intent in context. Individual user interfaces for specific functions are typically superior to combining user interfaces for different services. Opening a garage door has little in common with controlling a thermostat or turning off lights. Dedicated interfaces work best for these actions.
Home assistants in the cloud, connected through smart speakers, are making progress in their ability to interact with users. Still, they are usually confined to voice interactions, and are limited by the lack of context and history on which they can draw. While “skills” are being continually added, the ones that come together in “persistent” conversations involving prior memory are sorely lacking.
Automation of simple tasks is a major promise of the smart home. However, the current generation of devices and services has some way to go towards fulfilling that promise. Most automation offered relies on a single sensor stream or program for control. This provides rarely sufficient context for smooth operation of more complex functions. Every household has its own rhythm and ways of doing things. Personalization and consideration of context, be it from history or the current situation, improves the outcome of automation. In cases where automation succeeds only some of the time, what happens in the remainder of situations? If consumers have to override or even disable automation, it becomes an irritant.
The need to integrate multiple sensors and devices to achieve automation and to provide an overview of all IoT devices in a home has driven the emergence of smart home hubs. These hubs can bridge the various types of wireless network technologies and provide a single user interface for the connected sensors and devices. Many of them also interface with IFFTT, a web service supporting links between devices and many Internet services. In addition to the multifunction hubs, some manufacturers also require specialized hubs for their devices.
The lack of API standards makes it difficult to expose sophisticated function of connected devices through hubs compared to a dedicated interface on a smart phone app. Often, the functions of devices connected to a hub are completely unrelated to each other. For instance, the control of a garage door shares little with the control of lighting or indoor climate control. In these cases, connecting the devices to the hub provides little value, complicates the user interface, and creates additional security exposures.
Security of smart home devices is an ongoing concern that is unlikely to abate soon. A simple web search of “baby monitor security incident” returns many scary references. Professional analysis of weaknesses in smart home devices shows numerous risks. A common reason for this appears to be the battle between speed to market and due diligence. The frequent “bug fix” updates of smart phone apps are just one indicator of this losing battle. Another factor can be the attempt to provide a simple consumer experience, trading off ease of use against security. The great variety of hardware and software in home IoT devices exacerbates this problem. The more connected devices share a home network, the greater is the risk that the weakest device infects and exposes the others. Being located at the center of the smart home network, hubs represent a key exposure for spreading security risks from connected devices.
(Get the IBV’s reports for electronics executives: Security for IoT devices: Preparing for planet scale security: Getting it right in Electronics and for IIoT: Electronics industrial IoT cybersecurity: As strong as its weakest link)
In addition to security risks, smart home devices also pose serious privacy risks. An experiment at Princeton University shows that connected home devices transmit far more information into the cloud than needed for their operation. Given this state of smart home security and privacy, it will take some time for consumers to broadly adopt its services.
Having taken a good look at the current smart home market and its technologies, let’s move on to the next installment in this series to imagine the experience of smart living in a truly smart home.