If you think autonomous vehicles are complicated, you won’t believe what’s involved in making our entire world autonomous — from every door we walk through to every elevator we ride. The COVID-19 pandemic has brought the need for contactless technologies to the forefront of our minds speeding up the need for a world of autonomous things, and here’s what’s involved.
The COVID-19 pandemic has changed what we expect from technology innovation, driving demand for virtual connectivity and collaboration. New users of contactless technologies have learned to access services from home long before they would have otherwise adopted them. As we return to work and play in the real world, many of us now hesitate to touch everyday surfaces such as doorbells, elevator panels, light switches, and restaurant menus.
I foresee a business opportunity that might exceed the scope and value of the autonomous vehicle market as innovators race to create truly autonomous doors and a far larger universe of contactless interfaces for stores, schools, offices, and every other public place.
We can try to put some numbers to the opportunity, starting with elevators. There were an estimatedOpens a new window 900,000 elevators in operation around the world in 2018, and it is estimated that the number will increase by 70% to $135 billion in 2026. HotelsOpens a new window ? More than 700,000 around the world in 2016, comprising 15.5 million rooms.Â
Hospitals and healthcare facilities worldwide numberOpens a new window in the hundreds of thousands, with room counts in the millions. In the U.S. alone, there were overOpens a new window 160,000 gas stations and convenience stores in 2015, approximatelyOpens a new window 650,000 restaurants near the end of 2017 (let’s estimate a conservative half-dozen menus for each, or 3,900,000), and nearlyOpens a new window 140 million housing units in 2019.
Add the other locations like offices in multi-unit buildings, government facilities, transportation hubs, schools (and classrooms), and factories large and small, and then create a constant for an upgrade or replacement unit volume. The potential market size for such innovation is almost incalculably large.
Right now, delivering on that opportunity is still a vision.
Learn More: The Rising Importance of IoT Devices in Healthcare
Why Aren’t Automatic Doors Already Autonomous?
Automatic doors in public places are as ubiquitous as they are flawed. I’d bet most of us have had experiences in which doors opened repeatedly if you moved too close to an unseen sensor or, conversely, didn’t open even after you waved your arms frantically.
That’s because the use cases for creating a truly autonomous door are quite complicated. Does it open when someone stands in front of it, or does it require some level of security, and if so, what is it? Suppose it relies on a camera to sense motion or a sensor to receive voice commands. Can it differentiate between users (or luggage) and passers-by (or birds), or word choice and dialects, in all manners of ambient lighting and noise?
Perhaps most importantly, what is the default control if it fails in any of those cases? What’s the backup that keeps doors from refusing entry or closing on your nose before you step through them?
To summarize, truly automated contactless interactions must be robust, easy to maintain, and, to meet the enormous potential demand of the marketplace, should be able to operate with the same ease that we enjoy actuating tasks such as contactless payments from our smartphones.
Just like my door example, it won’t be good enough if restaurant menus more than occasionally yield incorrect orders, or elevators get us to our desired floors only some of the time.
Doors, menus, and elevator controls need to get smarter.
Learn More: Running a Database-as-a-Service in Kubernetes
The Algorithmic Challenge
The hardware for sensing and actuating data not only exists but has been proven in many ways in the automotive sector, where it has been put to use automating collections of tailpipe emissions, managing throttle position, monitoring pressures in the engine system, and gauging when braking needs to be applied automatically (ABS).Â
More recently, cameras and radar sensors have been put to use, providing data to drivers or prompting automated responses, such as lane position and backing up guidance. Over halfOpens a new window of all new cars sold in 2019 came with some type of voice recognition system.
While there have been notable examples of vehicles driving themselves (or nearly doing so) in certain environments — Tesla’s autopilot in the consumer space and Peleton’s commercial truck platooning examples come to mind — the developmental challenge remains with the software that informs those functions; the algorithms that interpret the sensor data, and make those robust and reliable decisions, don’t yet widely exist.
A door not recognizing an entrant obscured by lighting might not have the same safety implications as a car faces when it approaches a stop sign in bad weather, but the algorithmic challenge is similar.
The differences in these use cases also suggest intriguing development opportunities; for instance, the application of technologies that enable the recognition of hand or body gestures will be very meaningful in contactless controls. For instance, ultra-wideband (UWB) can provide precise GPS-like location data indoors as well as outdoors, and already comes pre-installed in the latest iPhones.Â
Another technology, broadly categorized as edge computing, promises affordable and robust local data processing at the immediate location where it’s then put to use (vs. relying solely on cloud connectivity), thereby enabling a generation of hardware and software tools that literally “rethink†how to deliver smart decision making.
Learn More: The Importance of Selling Support To End Users
The Way Forward Will Be Evolutionary
For the past decade or so, touchscreen development and implementation have been a business phenomenon, as the market grew up to ten times faster than overall displays (it almost doubled from $7.1 billion to $13.4 billion between 2010 and 2011), driven in many ways by the proliferation of smartphones and their uses.
But our recent experiences with COVID-19 suggest consumers will expect something differentOpens a new window . Further innovation in contactless interaction can not only deliver safer experiences but easier and more reliable ones, too.
The development path forward will be evolutionary, starting with solutions for simple use cases for proximity or gesture controls, such as allowing users to literally point at a choice among a limited number of options (like choosing a grade or type of fuel at a pump). Voice controls can work well in quiet or controlled environments but might need help from adjacent sensor technologies when there are too many voices or loud machines nearby.
The speed and depth of the development of both gesture and voice controls will largely depend on the innovation in machine learning (ML) and artificial intelligence (AI) that helps expand the range of choices and nuance of selections that can be made available to users.
The ultimate step will be creating wholly autonomous interfaces that both anticipate and then actuate user requirements, so think of elevators that “know†your desired floor from the meeting record in your calendar or the gas station that “remembers†your fuel type. This level of integration will require meaningful security protections, many of which already exist to enable mobile payments and maintain online data privacy.
The timing of these innovations will vary by use case and geography, but the development path is clear. The race is on to create a truly autonomous door and better contactless controls for just about everything else.
Did you enjoy reading this article? Let us know your thoughts on LinkedInOpens a new window , TwitterOpens a new window , or FacebookOpens a new window . We would love to hear from you!