Posted By Simon Montford on September 12, 2016
Next week I will be moderating "Interface Evaluation: Selecting the Optimum Interface for Smart Home Control" at the IoT Smart Summit, at 2.40-3.20pm on 21st September, with guest panellists: Diego Oliva (Founder, Glue), Joacim Westlund (CEO, FLIC), Kyrre Wathne (CTO, Viva Labs), Sam Woodward (Customer Education Leader, Lutron), Ricco Borring Winther (Director of Sales, Z-Wave Europe), and Philip Steele (Founder, nCube Smart Home).
For this reason, I decided to swot up on all the very latest trends in the smart home sector that specifically relate to human-machine interaction. One of the most obvious trends is devices that come with speech recognition software. Amazon's Alexa is currently king, well technically queen. This is because Amazon has done a superior job with information architecture, enabling it to respond to questions contextually, whereas "ask me anything" IAs (Intelligent Assistants) such as Siri, Now, and Cortana are far less proficient. Alexa has the advantage of being designed as a special-purpose device, so it will be interesting to see if Google Home, the much-hyped "Echo Killer", will be capable of deposing Queen Alexa from her throne. It is expected to be powered by Google Assistant, and will arrive in time for Christmas.
Developing speech recognition systems that are capable of humanlike conversation remains very challenging, which is perhaps why Apple has been so slow to launch its own "Echo Killer", and why other leading consumer electronics OEMs such as the LG SmartThinQ Hub are taking an "if you can't beat 'em, join 'em approach" by integrating Alexa into their products, instead of trying to compete head-to-head.
The two hardest nuts to crack are teaching human-machine conversational interfaces to understand context and all the subtle linguistic nuances that us humans take for granted. For example, take the word "hot". Its meaning can vary greatly according to context (temperature, spiciness, weather, attractiveness etc). Another problem with today's IAs are that they need to source data from multiple agents, making it hard for them to keep track when asked a series of related questions or a single complex question. This is why conversing with Siri and the like is disjointed and feels unnatural.
Not only is context awareness a huge challenge, but also the ability to program emotion. This is why Effective Computing is another hot trend in HCI. It looks like everyone is jumping on the "friendly home robot" bandwagon. Even Mark Zuckerburg revealed that he's been working on an IA to help him run his high-tech home. Other examples include Jibo, LG Rolling Bot, Asus Zenbo, Pepper, Sony Xperia Agent, Aido, Furo-i, Miro, Samsung Otto, Buddy, Big-i, and Tipron Cerevo.
In addition to voice command, there are several other ways humans and machines can interact. These include touch, gesture, and even thought! In my view the optimum interface combines the best of each. This is because the way a person chooses to interact is often down to subjective preference and circumstance. For example, some people may prefer touch because they find speaking to inanimate objects creepy, or perhaps they wish to enter private information such as passwords or financial details when others are within earshot. Gesture and/or voice will trump touch when the user is say wearing gloves, or undertaking tasks such as bathing or baking.
Touch will continue to be important, because humans are instinctively tactile. As infants, this is how we explore the world. This innate desire to touch, tap and swipe will therefore be one of the primary ways humans continue to interact with connected objects. For this reason we will continue to see touchscreen keyboards (Flexpansion, SwiftKey), remotes/home hub controllers (Neeo, Savant, Oomi, Wink,Tuxedo) and push button devices (Flic, Dash, Carling).
Incorporating motion controlled technology within the home is also gaining traction as motion sensors become cheaper and more accurate. Intel Curie is a button-sized module, which comes with an embedded motion sensor as standard. It is starting to appear in all kinds of wearables and smart home devices, other exciting gesture technologies to keep an eye on include Project Soli, Nod, Myo, and LeapMotion.
With regards to what the future smart home interface may look like, I predict over the next five years we will see more systems that utilise Augmented Reality. Data will be overlaid onto objects replacing the need for physical buttons, switches or OLED touch screens. Today's clunky headsets will become far smaller and will evolve into discrete wearables or embedded devices.
In ten years' time and beyond, we'll see Volumetric Display interfaces that will enable us to view 3D objects with the naked eye. Current prototypes aren't safe as they use lasers, capable of burning human skin, that are used to ionise air molecules. Japanese researchers have, however, found a way to make lasers safe by speeding up their pulse rate. Their new approach uses femtosecond 1Kh infrared pulsing lasers, which fire every millisecond and project holograms that are safe to touch. The same researchers believe that they will even be able to add a haptic dimension, so the hologram will feel like a physical object when touched!
Beyond ten years, we will all be at the mercy of robotic assistants like HAL 9000. I can imagine a scenario similar to the movie scene in "2001: A Space Odyssey" when HAL tries to kill Dave Bowman by refusing to open the pod bay door.
Joking aside, I predict that at some point Artificial Intelligence will be programmed to make decisions on behalf of us humans, based on (hopefully) our best interests. For example, an IA may intervene when domestic abuse is witnessed. During an altercation, it could attempt to protect the victim by denying access to certain rooms, or even the entire property, then attempt to calm the aggressor down while it calls for emergency assistance. Today this could be theoretically accomplished, using an if-this-then-that approach, for example, if lots of noise and movement is sensed that doesn't fit a usual pattern, then sound an alarm and call for help. What I'm suggesting is an IA that can genuinely think for itself, and respond autonomously.
Way before any of this sci-fi stuff becomes reality, we must first solve one of the biggest barriers to technological progress faced by the industry today - fragmentation. The smart home ecosystem is currently broken. Consumers require multiple networking devices, and applications to build and run their systems. There are too many connectivity standards, and too many companies that want to lock consumers into a single proprietary platform.
Those who have taken the plunge tend to be either the very wealthy or tech-savvy early-adopters, who are willing to risk their time and money on installing tech that may become obsolete. You only need to look at what happened to Revolv to understand the risk of potential obsolescence.
Another major issue is a lack of standardisation when it comes to UX. The smart home industry today is like the early personal computer era. The PC didn't go mainstream until end-users were given a user-friendly standardised operating system. Back then, it was Windows and Apple that dominated, but attempting to push closed standards in a post-Internet world no longer makes sense. Today the web browser is the operating system and IPv6 is the glue. I don't want to be forced into using proprietary apps to control my devices. Instead I want to purchase any device, and know I'll be able to control it via its IP address, and connect it to other devices. Oh and control where MY data is stored, and how MY data is shared. Is that too much to ask?
The following companies will be represented on the panel. If you have any questions, please share them below this post, contact me via Twitter @simonmontford or send them via email.
Glue replaces your physical key with an app. You can give out permanent, temporary or one-time digital keys in an instant and revoke access privileges as easily. Get instant notifications when someone unlocks or locks your door.
Flic is a simple and stylish wireless button that lets you create a shortcut to your favourite actions so that you don’t have to touch your phone.
Viva Labs offers a white label solution to "smart home integrators" that encompasses a learning thermostat, an autonomous alarm, and a home monitoring system; all powered by the company's proprietary AI. Based on the preferences and behaviour of home owners (in addition to sensor, GPS and weather data) their smart home assistant can learn how best to control heating, security, and lighting. Further products, such as smart locks, will be added to their platform but their primary USP is that everything is plug-and-play, so no complex setup and configuration are required.
Lutron Electronics designs and manufactures energy-saving lighting controls and lighting control systems for both residential and commercial applications. The company offers a wide variety of light dimmers, whole-home and whole-building dimming systems, controllable window treatments and colour-matched accessories, including wallplates, receptacles, fan controls and more.
Unlike other smart home hubs, Oomi Cube is not only packed with environmental sensors, a night vision IP camera, security detection and wireless control, but it is also the only one to feature a 360° IR blaster for home theatre and climate control. It offers more out of the box and doesn’t even need to be connected to the cloud to work.
Z-Wave is a wireless communications protocol for home automation. It is oriented to the residential control and automation market and is intended to provide a simple and reliable method to wirelessly control lighting, HVAC, security systems, home cinema, automated window treatments, swimming pool and spa controls, and garage and home access controls. There are hundreds of interoperable Z-Wave products marketed under different brands, and over 35 million have been sold since 2005. Z-Wave was developed by a Danish startup called Zen-Sys that was acquired by Sigma Designs in 2008.
nCube has put a brain at the heart of your home, linking smart tech together and making it easy to control. Unlike almost every system on the market, nCube’s brain works even when you’re offline. A unique feature of nCube is your information is never sent out to central servers or stored in the cloud. Instead it stays on your device in your home – meaning it’s much more secure and totally private.
© Simon Montford (WEB3IOT), 2014-2019. Unauthorized use and/or duplication of this material without express and written permission from this blog’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Simon Montford and WEB3IOT with appropriate and specific direction to the original content at web3iot.com.