Five disciplines for keeping a picture on air

Our relay team has fallen from fourteen engineers to nine, and the network still has thirty-nine relay stations on hills and in valleys. I plan maintenance around that fact. Every site now reports power, temperature and transmitter status remotely, so an engineer drives out because of a reading, not a calendar. Each site has a written access plan for winter, agreed with the municipality that clears the road, and a site record that captures what a retiring engineer knew. We also inspect two sites shared with another broadcaster on a reciprocal basis, which saves each company a winter journey. I do not pretend that monitoring replaces a visit. It decides which visits matter. The test I apply to any new equipment is whether the engineers we will have in five years can maintain it.

A feed to a relay station can travel over our own microwave link or a circuit leased from a carrier. I choose between them by asking what happens to viewers if the path fails, and for how long. Our backbone, which carries the signal to the main transmitter and the six largest relay stations, stays on microwave because it serves 86% of households and does not depend on another company's restoration priorities. Smaller spur feeds can use leased circuits, ideally over two physically separate routes. I learnt the importance of physical separation the hard way, when two carriers' cables shared the same poles for three kilometres. I now require route maps before signing, and an engineer walks any section where the routes appear to converge.

In March 2011 several of our small relay stations went silent when their batteries ran down. Since then I have treated backup power as a decision about which households keep a picture, and for how long. Twelve priority sites, covering most of our viewers, now have generators and fuel for seventy-two hours. Smaller sites have batteries for eight hours and a written order of refuelling visits that depends on road conditions. I tell the board those figures directly. I also run a winter exercise each year in which the master control team, the relay engineers and the newsroom rehearse a long power loss together, because the technical plan is useless if the newsroom does not know which areas will lose our signal first. Each exercise ends with a written list of what failed, and the list returns to me.

Our transmitters operate under radio station licences, and any change to a relay station involves the regulator's procedures, which can take most of a year. The Broadcast Act expects broadcasters to carry information that helps protect people during a disaster, and I treat that expectation as a design requirement for our network and our news stream. Leased circuits bring a different relationship. A carrier has its own obligations under the Telecommunications Business Act, but those concern its service, not our broadcast. Our contracts therefore set out restoration priorities, notification times and route information in writing. I rely on our legal adviser for interpretation and keep my own role to describing what the equipment does and what we have committed to. When the two do not match, I raise it with the president before the regulator or a viewer finds it.

Our revenue has fallen by almost a quarter since 2008, while transmitters, links and playout systems still reach the end of their supported lives on fixed schedules. I manage that by separating what must be renewed like-for-like from what can be shared, extended or retired. Some equipment can run safely beyond the maker's support with spare parts we hold; some cannot, and I say which. We now share inspections and spares with a neighbouring broadcaster. I present each renewal with its ten-year cost, including circuit fees and power, because an option that saves capital can cost more to run. I would rather show the board a near break-even than describe a change of method as a saving. Where a cheaper option shifts work onto engineers we cannot hire, I count that as a cost.