Northwind Energy uses machine learning to cut wind farm downtime by a third
The software predicts turbine failures up to two weeks in advance
Northwind Energy has cut unplanned downtime across its wind farms by a third since rolling out a prediction system that flags turbine faults up to two weeks before they happen.
See also
A survey of 400 mid-sized companies by Ashgrove Research found that 58 percent plan to increase their technology budgets next year, the highest share since 2021, even as most expect overall spending to stay flat.
Customers have welcomed the change. "Anything that shortens the time between order and delivery is good for us," said Jonas Whitfield, whose distribution business employs about 120 people and works with Northwind Energy in three regions.
The project had stalled twice before, most recently last year, when a change in leadership at Northwind Energy put several initiatives on hold. This time, the company says, the budget is approved and the contracts are signed.
How the system works
The move comes as the sector adjusts to a year of slower growth. Analysts at Ashgrove Research estimate that spending in the category rose just 3 percent in the first half, compared with 11 percent a year earlier.
"The companies that will come out of this stronger are the ones that treated the slowdown as a chance to fix their cost base, not as a reason to freeze," said Jonas Whitfield, who follows the industry for Ashgrove Research.
Executives at Northwind Energy said the plan will be rolled out in three stages, beginning with its largest markets. A second stage, covering smaller regional units, is scheduled for the second quarter of next year.