"I really would like my dog to do my groceries for me," says Brian Subirana. Sitting in the press room at Web Summit 2016, our chat probably isn't even the most far-fetched conversation in there. This friend to dogs spends his time as director of MIT's Auto-ID Lab - where the concept of the Internet of Things originated in 1999 - thinking about what's possible, what's close and what's "far, far away."
Subirana uses a framework to look at the question in terms of the information frontier (processing, storage, network), the matter frontier (breathable, high resolution management, legal) and the life frontier (sensing, human interface and artificial life). Here's what's coming down the pike.
It's an obvious question when talking to someone who sees cutting-edge tech first, all day everyday, but I ask it anyway. What has Subirana seen at MIT or elsewhere that's most exciting or unexpected?
The next level of stick-on is that you paint it on
"The smart skin we're doing in the lab is pretty unbelievable," he says. "We're working on humidity sensing and we have these paints that you can paint on your skin. Depending on the humidity, the material responds differently, like whether you're tired." I mention the trend for stick-on wearables that can measure health metrics and read facial expressions. L'Oréal's My UV patch [as seen in the top picture] and MIT's DuoSkin project with Microsoft Research also spring to mind.
"The next level of the stick-on is that you paint it on," Subirana counters. The lab's prototypes could eventually transform the beauty industry (and, I guess, make someone a boat load of money in the process). "Hopefully cosmetics will tell you when you need to put more or less cosmetics because of the humidity of the skin," he explains. "You could, for example, apply a regular cream then you have a stamp and the impurities you've created with the stamp are enough to measure the difference."
The Auto-ID Lab's prototypes of this technology are still early - at the moment it consists of a fixed copper coating. Alongside exploring how to learn more about our bodies, another problem that keeps MIT's scientists and researchers up at night is how to identify the bodies in a smart home.
"It's not me in particular that's working on it but MIT is doing some amazing work in Wi-Fi sensing," says Subirana. A couple of years ago, research into seeing through walls with Wi-Fi was announced - as our bodies reflect the signal - and it's still being investigated. As our smart homes get smarter, tracking us could be one of the last pieces of the puzzle.
"We now have these digital signal processing algorithms that can take a Wi-Fi signal and tell you who is on the other side of the wall," he says. "You could find out their heartbeat or breathing, where they are, whether they raise their arm, if they move. The same digital signal processing allows you to uniquely identify a chip [i.e in a smart home device]."
The timeframe for breakthroughs in locating us in the smart home is in the next five years, he believes. But right now, we're limited: "It's impossible for a device, a phone or a watch, to know how close you are to a coffee machine." The implications for this are pretty huge in terms of the AI-powered smart home interacting not just amongst itself but with the humans it is designed to help.
Robo waiters and delivery dogs
Moving further into the realm of what's possible, there are a couple of particular challenges of interest to MIT's various labs and staff that affect both the inside of the smart home and business logistics. The first is the "Amazon challenge" of building robots that can grasp and grab objects - from a shelf in an Amazon warehouse, for instance, but also "an arbitrary object from the supermarket and putting it onto a cart or picking up a key and putting it into a door. We're a long, long way from opening the door." Or even better, "imagine if a robot can go pick up a cup of coffee and give it to you like waiters." Cut to slide of the flabby hovering humans in Pixar's Wall-E.
New Deal Design's Whistle smart dog collar
And for fans of pet wearables, geo-fencing could in theory turn your dog into a delivery collection pooch. Subirana likes to illustrate how he views progress and obstacles for the Internet of Things with this genius example: "You have a standard pet wearable so that they can walk through different neighbourhoods," he says. "Each neighbourhood would have its own policy for walk-along dogs, each dog would have its own certificate. They will be much more healthy and they will be able to go and visit your Grandma if she's alone."
There's more. "Dogs are ecological for last mile logistics which is a major problem.. You can take a package that you want to send to a friend. You can just put it out there, maybe a dog will pick it up, bring it to a post office. You could imagine dogs then waiting for an Uber to stop, the trunk opens automatically, throws the package, drives away. I would like taxi cabs to be able to use their empty trunk space."
Bringing us back to reality, the Auto-ID Lab director signals that technologies such as augmented (or mixed) reality and Bluetooth - both fairly omnipresent at this year's Web Summit - still have a ways to go.
"I was a bit disappointed with Facebook's CTO. He said [at the conference] 'This is the year of augmented reality'. It's one of those things that big companies say but they don't provide any evidence," says Subirana. "What we're doing is we have some research projects on shopping. So we think that if you go into the grocery store and you have glasses that help you shop, maybe that's a use case. So we're working hard on that, there may be something there."
And of course sports, though again he's not yet 100% sold: "Augmented reality may be good in stadiums. When there's a penalty kick by Messi - you can put on your goggles and be the goalie and be Messi. Maybe, maybe, maybe..." As for Bluetooth, "they need to come up with Bluetooth 10.0," he says, "they need to improve the security and automatic connectivity. Even when you connect your car, it's very painful."
If you feared the Internet of Things is reaching a plateau, it seems there's no need to fret. A couple more throwaway ideas to leave you with: an Amazon Dash-style smart home camera for instant ordering, and connected dogs fighting Zika by dispersing artificially grown mosquitoes.
How we test