#forecast

2025 review — Edge drivers local, robots choreographed

“Seven and a half of eight on last year's forecasts — but the most useful line is the one I got wrong twice and finally reversed: the solar math, which I'd said no to three years running, flipped to yes.”

2025 pet IoT — the DIY era, properly arrived

“The same year, the same small company inverted two of my calls. Tractive shipped the resting-vitals collar a decade of giants never managed — then bought Whistle from the giant that gave up. The empire didn't expand; it sold its device leg to the underdog that had just out-engineered it.”

2024 review — Bespoke kitchen, Frame TV, Matter bridges

“6/8 on last year's forecasts. The two I missed were both AI-on-camera bets — and that's the honest pattern: the hardware lands on schedule, the intelligence layer keeps slipping a year.”

2024 pet IoT — Atom passed, Find My trackers landed

“Looking back: a hard year. The pet-IoT data did its job — caught Atom's decline if I'd been looking at the right metrics. The lesson, expensive: dashboards must adapt to the pet's life stage. I missed the signal because I wasn't watching the right view.”

2023 review — built the house, kitchen has an OS

“Looking back: 6.5/8 on 2022's forecasts, ~85%. The new house transformed the smart home from 'a layer on top of a house' to 'the house was designed for this.' Different rules apply now.”

2023 in pet IoT — a new house, and the year the data started to matter

“For ten years this was a hobby — gadgets on healthy animals, data I collected because I could. This was the year it stopped being decorative: Atom's collar flagged a real change before I'd admit I was seeing it, and the vet found a heart murmur. The category is finally being tested for what it's actually for.”

2022 review — Matter shipped, the LoRa garden worked

“5.5 of 8 on last year's forecast — about 70%. Matter hit on schedule, ESPHome kept exploding, HA Yellow shipped. The honest miss: I scored myself a Powerwall and then talked myself out of it.”

2022 in pet IoT — Quark arrived, Halo went back, the cat side finally got smart

“The household hit four animals this year, and for the first time the cat side got the analytics layer the dog side has had since 2013. Nine years to instrument the cats. The dog side was the easy half all along.”

2021 in review — cameras everywhere, Coral landed. 2022 forecast.

“Five of eight on 2020's forecasts, ~69%. Cameras, ESP DIY, and the garage controller all hit; the misses were timing — cross-camera tracking and Wink's actual death are both still 'next year,' and Matter is still a spec without a single product.”

2021 in pet IoT — AirTag landed, Whistle started reading behavior, Halo I won't buy

“AirTag was the most-hyped pet tracker of the year and I don't have one on a dog — because the same anti-stalking chirp that makes it safe for people makes it useless for a pet that can't tell you it's beeping.”

2020 review — local-first arrived early, alarm landed

“Looking back: ~7/8 on 2019's forecasts, the best year since I started scoring. Smart lock and the Coral hit; Wink pivoted to a ransom-note subscription instead of folding; the Z-Wave rewrite is right on the doorstep but not quite shipped. The pandemic compressed two years of smart-home maturation into nine months.”

2020 pet IoT — Petnet died, Boson arrived, multi-cat real

“Four of eight on last year's forecasts, and the misses all rhyme: every 'health' collar still infers from behavior, no litter box knows which cat, and I keep building the smart part myself. The category surges; the intelligence lags.”

2019 review — local-first arrives, security maturing

“Looking back: 5.5/8 on 2018's forecasts. ~73%. The miss was Frigate's maturity — running on a Pi without a Coral, the detection was too slow.”

2019 pet IoT — Fi shipped, Whistle went health, Petnet hung on

“Seven years in, the lesson I keep relearning is that the device almost never fails — the company behind it does. Fi's battery was the engineering win of the year. Petnet still being alive was the surprise.”

2018 review — quiet year, Zigbee goes local, the doorbell got scary

“A quiet year on the blog was a busy year in the closet. Three posts, but the architecture moved further toward local-first in 2018 than in the three years before it combined.”

2018 review — Lovelace, leak shutoff, smoke + CO

“Looking back: 6.5/8 on 2017's forecasts. ~81%. Better than 2016 (64%) and 2015 (71%), worse than 2013 (83%). The arc seems to be: my confidence calibration is improving; my Apple-timing intuition is not.”

2018 pet IoT — the quiet year before Fi

“The interesting thing about 2018 wasn't a product. It was the absence of one — Atom went four months bare-collar because nothing on the market was worth buying, and I'd rather wait for Fi than pay Mars again.”

2017 review — Home Assistant lands, cheap Zigbee opens

“Looking back: 5/8 on 2016's forecasts. ~64% — my worst year so far. Two of the misses were timing (Apple HomePod slipped, Echo Show shipped earlier than I expected). The big call I got right: Home Assistant is the unifier I'd been waiting for since 2013.”

2017 review — Home Assistant arrives, and the journal goes quiet

“Local-first isn't a philosophy until the day the cloud you depend on has an outage and your lights still work. 2017 was the year I stopped trusting someone else's servers to run my house.”

2017 pet IoT in review — the cat side finally gets data, the feeder gets scary

“Five years in and the misses are always the same shape: vendor ship dates. The structural calls keep landing; the calendar calls keep slipping. I predict on a calendar; vendors don't ship on one.”

2016 pet IoT in review — Whistle goes cellular, Mars buys in, Furbo lands

“The big call I got right was Whistle going cellular and getting acquired in the same breath. The miss, again, was timing — I keep forecasting on calendars, and vendors ship when vendors ship.”

2016 review — Google Home, HomeKit, Hue automation

“Grading the 2015 forecast honestly lands me at 69%, not the 88% I'd hoped — and the miss is the same one every year: Apple's rumored smart speaker that keeps not shipping. The pattern is that I forecast Apple on a calendar, and Apple ships when Apple ships.”

2015 review — voice + HomeKit land. Security starts

“Looking back at the 2014 forecast: 5/7 hit, 2 missed. ~71% accurate. Worse than 2013's 83% — partly because I was more confident this year, partly because Apple and Google moved at different speeds than I predicted.”

2015 in pet IoT — the cat side finally shipped, the dog side split in two

“The cat side of the house went from one passive chip to a door that reads it six hundred times a month. That's the whole story of 2015: the most analog member of the household finally got a radio pointed at her.”

2014 review — SmartThings + Alexa change everything

“Looking back at the 2013 forecast: hub-as-unifier (✓), Amazon voice (✓), HomeKit announcement (✓), Zigbee 3.0 (just-barely ✓). Net forecast accuracy: 83%.”

2014 in pet IoT — the dog side grew up, the cat side never started

“First full year of pet IoT in the house: the dog has a tracker, a market, and a year of data I'd actually act on. The cat has a chip the size of a grain of rice and nothing else. That gap is the whole story of the year.”

2013 review — three vendors, one annoyed homeowner

“Voice control is the inflection point. Whoever wins voice in 2014 owns the smart-home conversation for the next decade.”