A visit to an Internet of Things conference at MIT shook my faith in people's ability to design and maintain complex systems and navigate them intelligently.
The morning went sharply downhill when I pulled into the exit of the confusingly marked Alewife public transit station in North Cambridge, Mass., only to find the garage full. An awkward wrong-way drive toward the exit drew an annoyed look and reprimand from a policeman. I had no choice but to venture into the bowels of Kendall Square. My Apple iPhone's Google Maps would tell me where to go.
After a slow crawl through tony neighborhoods, I was quickly wending my way east on Memorial Drive, the boulevard between the Charles River and the Harvard and MIT campuses. But then the scene before me caused me (mistakenly, it turned out) to distrust Google Maps. Surely it wasn't saying to go straight onto that narrow, driveway-looking thing, my view of it partly obscured by snowbanks from the historic winter. A right turn was the only legal option -- across the Massachusetts Avenue Bridge over the river into the tangled old cow paths of Boston.
But Google Maps didn't know the nearby Longfellow Bridge had long been closed for construction and kept telling me to take it. My wetware database knew, but the query had gone out too late. Now, in that typically Boston way, I found myself essentially forced onto the wrong highway ramp, speeding away on Interstate 93 toward home.
A slow, circuitous route to reverse direction -- wetware and Google Maps were in sync for once -- brought me back to Boston. Then the phone died and my Internet thing was no longer connected.
So, I was forced to rely on the aging default system, its simple heuristic guiding the way. Take the next bridge (Mass. Ave. again) across the river and go to the first big clump of buildings, since MIT comes before Harvard. Drive until you find an open garage (on-street parking is almost nonexistent) and don't worry about getting close because in this neighborhood, driving takes longer than walking.
Now I was officially late, but finally parked. Then, an unfamiliar user interface: a stairwell with locked doors, and a wall phone for calling security to let you out. I told them I was on the fifth floor but forgot I had moved a couple of floors. Sweating up and down the stairs, I could hear us missing each other while I tried other locked doors like a mouse repeatedly picking the dead ends in a maze.
Eventually I was outside and facing the famous MIT Media Lab, which ever since its mid-1980s founding I had revered as a liberal arts major's Promised Land from the insular world of hobbyists, programmers and DOS command lines that personal computing was back then.
Inside MIT Media Lab, "future-obsessed product designers, nanotechnologists, data-visualization experts, industry researchers, and pioneers of computer interfaces work side by side to tirelessly invent -- and reinvent -- how humans experience, and can be aided by, technology," says the mission statement.
Products as the new platform
As if infected with the spirit of the place, MIT Connected Things 2015 was full of very smart people who are idealistic about the potential of the IoT and working hard on it, but also refreshingly honest about the huge technical, financial and strategic hurdles.
Panelists on the industrial IoT track presented the same balanced view, but with the benefit of laboring in sectors such as manufacturing and the supply chain that have a longer, deeper history of connecting things than do retail and healthcare, the other conference tracks. Machine-to-machine, or M2M, technology was arguably the industrial IoT's immediate predecessor, and programmable logic controllers and RFID sensors have been capturing data for decades. Telemetry is the old name for remote monitoring of sensors.
But these industrial IoT advocates said there is much to be gained from putting the old systems on the Internet, and retrofit kits could be an IoT growth industry. Most of them touted the potential of treating their products as platforms for delivering lucrative services to consumers who are using the IoT to become more engaged in managing their relationships with vendors, products and services.
"Product as a service" was on the lips of several speakers, including closing fireside-chat guest Jim Heppelmann, CEO of PTC, the vendor of computer-aided design software that has branched into product lifecycle software and is betting big on IoT.
"We were helping to create mechanical things," he said. "At some point, we realized they weren't just mechanical anymore. Then we got into helping people service things." Now manufacturers will have to adopt a "product as a platform" mind-set to ride the IoT.
The drive toward IoT standards and interoperability testing, while still urgent, seemed less daunting after hearing from industrial experts who have already built applications under a different name or can see a clear path to the next level. Exploiting "small data" is the true practical challenge, according to several.
Perhaps Forrester analyst Frank Gillett captured the situation best when he said, "It's a mistake to think of it as one monolithic thing, like cloud. It's a use case across verticals rather than a top-down thing. This is happening bottom-up."
The future of the Internet of Things in a trash can
One example -- admittedly one of the conference's few real-world cases -- came from Brian Phillips, executive vice president at Bigbelly, which makes solar-powered sidewalk waste and composting bins that have sensors connected to management consoles over the cloud. Cities including Los Angeles and campuses such as MIT itself use the global network to monitor when bins are full, reducing pickups by up to 80% and running analytics on local recycling trends. "We send up a kilobyte of data, at most, for each station," he said. Now Bigbelly is moving to outfit bins to augment public Wi-Fi and provide "urban intel" on metrics such as pedestrian traffic.
On my way out, I stopped to visit an exhibit that pays homage to the Media Lab's visionary founders, including Seymour Papert, who invented the Logo language to teach children how to program by controlling the simple movements of a robotic "turtle."
Still drawing from wetware reserves, I found my way home quickly and uneventfully.
Learn about IoT challenges
Read a column about Internet of Things hype
Understand IoT infrastructure issues