05 · Nokia
Black Summer, in order
Bridges through the 2019–20 bushfires with the RFS and the Telco Authority. Offshore Tier 1 could declare a major without waiting for Sydney.
RFSRestoration order
The 2019–20 Black Summer fires were not a single incident with a start and a restore. Sites failed in the order the fire chose, often faster than a Sydney morning could be assembled, and the people who needed the network back first were not always the people shouting loudest on the bridge. I ran those bridges with the Rural Fire Service and the Telco Authority.
What we agreed, and then enforced on every handover, was a restoration order. Critical public sites came back before commercial volume. Field, the operator and the offshore teams worked from that list. A new shift was not allowed to invent its own priority because the previous shift had gone home.
The second failure was quieter and just as expensive. Offshore Tier 1 could already see a fault that met the threshold for a major incident, and they waited for Sydney to say the word. That wait is not caution. It is a defect in the operating model. I wrote a triage they could apply on their own shift: the symptom, the number of sites, and a public-safety flag. If it matched, they declared. They did not wait for me to wake up.
I still took direct command of every P0. Declaration and command are different jobs. Making the declare possible offshore, and then coaching the commanders — including nine I took through in person in India during a structural change — is what kept the rule alive after the season, instead of leaving it as something I remembered.
How I ran it
The order of decisions, written out so the clock is visible.
- 01
I treated the season as the event. A single fallen site was handled as a site. The command problem was the order in which many sites would be restored while the fires were still moving.
- 02
With the Rural Fire Service and the Telco Authority we agreed the list in plain language: critical public sites first, then the rest. Complaint volume did not outrank that list, including when a large customer was loud.
- 03
One bridge held field, the operator and offshore. At every handover I restated the priority so a fresh shift could not quietly replace it with whatever was paging hardest.
- 04
The delay I went looking for was not in the radio. Offshore Tier 1 could see the pattern and still waited for Sydney to declare a major. That gap sits between detection and command, and it costs the sites that are already down.
- 05
The triage was written so they could use it without me. Symptom, site count, public-safety flag. A match meant they declared on the shift. A Sydney morning was no longer part of the path.
- 06
Every P0 still came to me. They owned the decision to declare. I owned the worst of the bridges. The nine incident commanders I coached in India were how that split survived a night shift and a business change, rather than depending on one person in Sydney.
- 07
After the season the rule stayed in the process. A bushfire triage that lives in the incident controller’s head is not a triage. It is a memory, and memories do not roster.