The stated problem in 2015 was that the Council chambers ran a multitude of equipment and software across three separate production platforms, and the City wanted a single-source replacement.
Four requisitions, eight years, each picking up where the last stopped
A 2015 assessment and proof of concept established the problem in the organization’s own words. A 2018 replacement phase was carried alongside the Emergency Operations Center extension, producing concept drawings, capital improvement detail and the budget request. The 2019 to 2021 phase went through draft solicitation review, a pre-proposal meeting, an evaluator group, award and construction — construction in the literal sense. A second phase followed in the next fiscal year across two further committee rooms and the control programming.
Eight years and four requisitions is what one chamber rebuild costs an organization that can spend only what a council appropriates to it. The architecture has to survive being funded in pieces by the body it serves.
Decoupling the vote from the broadcast
The voting path was separated from public streaming onto independent encoders, so that a fault in the broadcast path could not reach a roll call.
A verification script was specified to be run against the voting system before a legislative session opened. A control-panel function was added letting chamber staff move a member onto a backup machine without interrupting proceedings. Relocating the voting platform onto a virtual machine inside the City data center was evaluated and recommended.
And the primary voting device was set to be the machine at the member’s desk, with the tablet as backup rather than the reverse — because tablet problems were few and usually cleared quickly. That is prioritizing by the consequence of a failure rather than by its frequency, which is the harder way round and the right one when the thing that stops is a vote.
Holding the change rate down
With multiple changes running simultaneously across the chamber estate, isolating a cause becomes impossible. The recommendation on the record was to make one change at a time.
Staffing was committed alongside the method rather than instead of it: two resources on named meeting dates, operating behind the scenes while monitoring the systems and supporting the room.
Slowing a program down in front of a customer who wants movement costs something in the short run, and it is the only way a cause is ever found.
The procurement route and the acceptance standard as part of the architecture
The program was driven hard from the Council Secretary’s office, and the governance recap assigned obligations by role. Three possible routes existed for a sole-source waiver, and nothing moved until somebody chose one — which is what public procurement actually looks like from inside.
The acceptance standard then set cutover around the Council’s calendar rather than the project’s: install during the winter recess, then test the backbone and every system with live proceedings. There is no test environment that contains a quorum, so acceptance proven in a live session is the only acceptance that means anything.
The Emergency Operations Center extension, and a question about power
The Emergency Operations Center extension into an adjacent chamber was demonstrated as an interim capability in 2016, broadcasting information disseminated from the Emergency Operations Center. The physical conversion was costed to repurpose Council’s ceiling lifts and monitors into that room so it could carry both missions.
Alongside it, a question about a proposed new first-floor hearing room: whether it should be tied to the building generators, because the main Council chambers are not generator-backed.
The resilience question about a council chamber is not the equipment. It is whether the room has power on the day the city has to meet.
Outcome
A chamber where a broadcast fault can no longer reach a roll call, accepted by testing in a live session rather than on a bench.