An Ageing Estate with One Unreliable Lift

The main opportunity cost of Option A is the package of risk-reduction measures that cannot be funded with the remaining HK$0.8 million. Full lift replacement costs HK$7.2 million, leaving too little for the HK$2.1 million smart monitoring system, the HK$1.8 million nurse service, or even the HK$1.1 million temporary external lift. The fire services requirement that the lift remain usable during works makes this acute: without the temporary lift, the 10-month works may strand residents on upper floors, and the elderly bear that cost. The benefit of Option A is a permanent fix, which matters because the lift has failed four times in six months. But the opportunity cost is not just money; it is the foregone chance to reduce daily fall risk and isolation for 41% elderly residents during the works. The corporation should therefore compare A not with 'doing nothing' but with a mixed package that keeps the building usable.

Recommendation: fund Option B, Option C, and the temporary external lift, and use the remainder for a lift-maintenance contract and a funding application for replacement. B (HK$2.1m) + C (HK$1.8m) + temporary lift (HK$1.1m) = HK$5.0m, leaving HK$3.0m. That is not enough for the HK$7.2m replacement, but it can pay for a 12-month maintenance contract, a spare-parts reserve, and design work for a future replacement. This package addresses immediate safety and isolation: the call centre and fall sensors help when the lift fails, the nurse visits the most frail residents, and the temporary lift satisfies the fire requirement during any works. Option A alone is non-compliant because the remaining HK$0.8m cannot cover the temporary lift. The trade-off is that the root cause - an unreliable lift - remains; the corporation should set a deadline to secure additional funding. A quick 'win' before an election is not a sufficient reason to skip that plan.

Privacy objections should be met with consent, data minimisation, and a non-surveillance design. The system should be opt-in, with a written explanation that it detects falls and inactivity, not movement or conversation; no audio or video should be sent to the call centre. Data should be processed locally where possible and retained for no longer than necessary, with a clear access log. Residents who refuse should still get a pendant alarm or a daily phone check, so refusal does not mean abandonment. The corporation should run briefings in Cantonese and other languages, use familiar faces such as the community nurse, and let residents test the device before signing. A 24-hour call centre is only useful if residents trust it; if they see it as 'being watched', they may disable it or not answer. Privacy is not an obstacle to care here; it is part of whether the care is accepted.