Toronto Quayside: 1,524 Pages, 190 Acres, Zero Sensors [2026]
Project Smart City · 01 · Project in Focus
Toronto Quayside: 1,524 Pages, 190 Acres, Zero Sensors
The most-funded smart-city project in North America never got to the physics. It broke one layer up.
Written by Scot Free
The series opener ran the audit on a parking puck that leaked. This one runs it on the opposite kind of failure. Nothing at Quayside ever leaked, because nothing at Quayside was ever installed. The project that was supposed to be the reference deployment for the connected city produced a master plan, a governance fight, a cancellation, and, six years later, a housing development that does not use the words "smart city" anywhere in its announcement.
That makes it the cleanest case in the series for a question the vendors never ask: what fraction of smart-city projects die before the pipe is even built?
The claim
October 2017. Sidewalk Labs, the urban-innovation company Alphabet launched in 2015, wins a Waterfront Toronto bid to redevelop a 12-acre parcel on Lake Ontario called Quayside. Alphabet pledges US$50 million for the planning phase. The pitch is a neighborhood built "from the internet up": sensor-timed traffic signals, dynamic curbs, underground freight tunnels, tall-timber towers, and a data layer under all of it that would make the district a testbed for how cities should run.
June 2019. The Master Innovation and Development Plan arrives: three volumes, 1,524 pages. It estimates the build at C$3.9 billion, with about C$900 million from Sidewalk Labs. It also asks for something the original bid did not: a 190-acre "IDEA District" covering roughly a third of the eastern waterfront, of which Sidewalk would lead development on 16 percent, plus an independent "Urban Data Trust" to govern everything the sensors collected.
Those are the invoice lines: 12 acres that became 190, a US$50 million commitment that became a C$3.9 billion plan, and a data-governance body designed by the company that would generate the data.
The pipe
On paper, the best one ever proposed for a district this size. The plan described sensors on pedestrian movement, traffic, weather, and building systems, a standard open mount for installing them, digital infrastructure shared across the district, and a data layer that was meant to be published as a public asset with royalties flowing back from technologies built on it. If you read the technical sections of the MIDP as an engineer, most of it holds up. The radio, the enclosures, the ingestion layer, the standards: none of it was the problem.
The pipe had one component the plan did not treat as engineering, and it turned out to be the load-bearing one. Who owns the data, where it lives, and whether it is de-identified at the sensor or somewhere downstream. That is a governance question, and in a public district it is also a physics question, because it decides whether the sensor gets installed at all.
What the data shows
There is no telemetry from Quayside, because there were no sensors. The record is the paper trail, and it is unusually complete.
| Date | What happened | Layer that moved |
|---|---|---|
| Oct 2017 | Sidewalk Labs wins the Quayside bid. US$50M planning pledge. Sidewalk's CEO says publicly it is not yet clear who would own the data. | Governance, undefined |
| Oct 2018 | Ann Cavoukian, Ontario's former privacy commissioner and Sidewalk's paid privacy adviser, learns at an advisory panel meeting that de-identification at the point of collection will be encouraged for third parties, not required. She resigns the next day. Waterfront Toronto board members also leave. | Governance, contested |
| Nov 2018 | The Intercept reports Sidewalk has set aside US$11M of the US$50M for public engagement. Opposition group #BlockSidewalk organizes. | Budget, reallocated to persuasion |
| Dec 2018 | Ontario's Auditor General publishes a critical report on the project's procurement and oversight. | Procurement |
| Jun 2019 | The 1,524-page MIDP proposes 190 acres and the Urban Data Trust. Waterfront Toronto's chair calls the expansion premature in an open letter and lists threshold issues that must be resolved before evaluation begins. | Scope and governance |
| Oct 31, 2019 | Sidewalk concedes every threshold issue: back to 12 acres, the Urban Data Trust and the term "urban data" eliminated, all personal information stored in Canada, data treated as a public asset under existing regulators, and Sidewalk will not be lead developer. Waterfront's chair: "This is strictly the Quayside project. Period. End of story." | Every non-physics layer, reset |
| May 7, 2020 | Sidewalk Labs withdraws, citing economic uncertainty and real-estate conditions during the pandemic. The Globe and Mail's post-mortem headline says the downfall began long before COVID-19. | Exit |
| Feb 2022 – Dec 2022 | Waterfront Toronto selects Dream Unlimited and Great Gulf. Agreement approved: 800+ affordable units, 3.5 to 5 acres of public space, Canada's largest tall-timber structure, an all-electric zero-carbon district. | Rebid as housing |
| 2025 – 2026 | Financing announced for roughly 1,200 rental apartments in phase one. Construction targeted for 2026, delivery 2030–31. Mix shifted toward purpose-built and affordable rental "in response to market conditions." | Build, nine years after the bid |
Run the reconciliation on that table and the variance is total. Sensor networks proposed: one district's worth. Sensors installed: none. Land requested: 190 acres. Land built: 12, by someone else. Data governance body proposed: one. Data governance body created: none, because the existing regulators kept the job. The one line that survived from the 2019 plan into the 2022 agreement is tall timber, and that is a construction method, not a smart-city technology.
Where it breaks
Read the "layer that moved" column top to bottom. Not one entry is a sensor, a radio, a battery, or a data pipeline. Every break is one layer up: who owns the data, who owns the land, who leads the build, who approves the technology. The physics was never tested because the project never reached it.
Three specific failure modes, in order of when they fired:
- Governance left undefined at signing. The CEO said in 2017 that data ownership was not yet clear. That is an honest answer and a fatal one, because in a public district the data question is not a detail to resolve later. It is the procurement. Twelve months later the privacy adviser resigned over exactly that gap, and the project spent the next eighteen months negotiating what should have been in the original agreement.
- Scope creep against a public counterparty. Going from 12 acres to 190 between the bid and the plan would be aggressive with a private landowner. With a tri-government agency, after a leaked document had already floated 350 acres, it converted a development partner into an adversary. The October 2019 concessions were sweeping and they came too late to rebuild trust.
- Persuasion budgeted where engineering should have been. US$11 million of a US$50 million planning budget aimed at public engagement is a signal. When a fifth of the money is going into the argument, the argument is the project.
The stated cause of death, pandemic economics, is real and also not the finding. A project with a signed data-governance framework and a 12-acre scope in 2017 would have been under construction by March 2020, and pandemic economics would have been a construction-financing problem, not an exit.
The Underground Take
The opener's rule was: ask to see the enclosure. Quayside adds the rule that comes before it: ask who signed the data agreement, and read the date. If the answer is "we'll work that out," you are not looking at an engineering project yet. You are looking at a negotiation with a sensor plan attached.
I ran nine connected-device programs for a Fortune 500 wireless bet, and when I sorted them by what killed them, three died on physics: firmware, battery, stability. The other six never got that far. They died in a proposal, a product requirements document, a dev-kit review, or a legal department. One of them had a signed master services agreement from a city in hand and still died in-house. Quayside is that pattern at municipal scale, and it is the pattern this series expects to find most often. The pipe fails in the field. Smart-city projects fail before the field.
The honest ending is that the 12 acres are getting built, as housing, by a developer that never promised a sensor. Torontonians are getting 800 affordable units and a timber tower. That is not nothing. It is also not what 1,524 pages described, and the variance between the two is the number nobody put on the slide.
Sources
- Government Technology, "Citing COVID-19, Sidewalk Labs Abandons Waterfront Toronto," May 2020 (US$50M pledge, 1,524-page MIDP, C$3.9B estimate, C$900M Sidewalk share, technology list).
- CBC News, "'Not good enough': Toronto privacy expert resigns from Sidewalk Labs over data concerns," Oct. 2018. StateScoop, "Digital privacy concerns will follow Sidewalk Labs to next venture," June 4, 2020 (Cavoukian's account of the October 2018 advisory panel meeting).
- The Intercept, Nov. 13, 2018 (US$11M of the US$50M allocated to public engagement; Cavoukian resignation letter).
- The Globe and Mail, "Sidewalk's end: How the downfall of a Toronto 'smart city' plan began long before COVID-19," May 24, 2020 (Auditor-General report, Dec. 2018; adviser departures).
- Waterfront Toronto, "Open Letter from Waterfront Toronto Board Chair," Oct. 31, 2019, and "Overview of Realignment of MIDP Threshold Issue Resolution" (190 acres to 12; Urban Data Trust eliminated; data stored in Canada; lead-developer change). The Logic, "Breaking down the deal between Sidewalk Labs and Waterfront Toronto," Oct.–Nov. 2019. BetaKit (350-acre leaked figure). Dezeen, Nov. 1, 2019.
- StateScoop, "Sidewalk Labs pulls out of Toronto Quayside project," May 7, 2020. Wikipedia, "Quayside, Toronto" and "Sidewalk Toronto" (2017 data-ownership statement; May 7, 2020 withdrawal).
- Waterfront Toronto, "Dream Unlimited and Great Gulf Group selected to develop Quayside," Feb. 15, 2022; Great Gulf, "Waterfront Toronto and Quayside Impact confirm agreement," Dec. 12, 2022 (800+ affordable units, public space, tall timber, construction toward 2026). City of Toronto Executive Committee materials, 2025; UrbanToronto reporting on Dream Impact phase-one financing (~1,200 rental homes; delivery 2030–31).
- The nine-program count in The Underground Take is the author's own portfolio; the programs are described generically and will be the subject of a separate post in this series.
Corrections and additions: Scott@IoTunderground.com. The community is invited to extend or correct any finding in this series.
