This week, the IGI volunteers are running the last of their validation studies. That involves hitting the same limits of detection 19 times out of 20, and reproducing positive and negative results produced at other labs. While they don’t have to wait for the FDA to give them a greenlight–labs have up to 15 days to submit their validation data for approval and can technically begin testing patient samples in the meantime–IGI has opted to wait until the review proves their tests work well. “Since we’re new at this we don’t want to be in a position where we have to go back and tell patients their results were wrong,” says Lin Shiao. Once they get the go-ahead, these volunteers will work in three teams to cover two 5-hour daily shifts, with socially distanced “battle lieutenants” that can step up if anyone falls ill.

The IGI testing rollout, for now, has limits. They plan to begin on Monday, running a few hundred tests per day, with teams running manual protocols on two of the older PCR machines. Later in the month, once the robots are fully validated, they expect to ramp up to as many as 4,000 daily tests, as needed, says Urnov. To start, only UC Berkeley staff and students will be eligible for testing, while the administrators work to get clearance to start accepting samples from hospitals elsewhere in the East Bay. “We would like to accept community samples,” Schaletzky says. “That was the whole goal from the start.”

Other Bay Area medical centers offer high-throughput testing. UCSF, for example, can now process 400 tests a day, says Bob Wachter, chair of the Department of Medicine there, which is enough to meet the health system’s current clinical needs. That has allowed the UCSF testing facilities to begin taking on tests from regional care providers that don’t have their own testing capabilities. But most other hospitals are stuck sending samples off to commercial or state labs with a four- to five-day turnaround.

Waiting a few days can be a problem, Wachter says. While doctors wait for tests to come back, they’re often forced to treat any people with respiratory issues as potential Covid-19 patients, just to be safe. That means assigning them to an increasingly short supply of isolation rooms and requiring any health care workers who interact with them to don masks, glasses, gloves, and other increasingly scarce personal protective equipment, or PPE. “It’s not that they’re not getting the right treatment, but they’re taking up beds that we might need,” he says. “The majority when they come back are negative.”

Kris Kury, an emergency room pulmonologist and medical director at Alta Bates Summit Medical Center in Oakland, told WIRED that being able to rule out patients more quickly would help hospitals better manage their supplies of protective gear like masks and gloves ahead of a surge in Covid-19 patients. For now, coronavirus-positive patients still make up a minority of people she sees coming in with respiratory symptoms. But until the tests come back, she and other health care workers have to treat them like Covid-19 cases and don protective gear every time they interact with them. “You can’t pull people out of isolation until you know they are negative,” says Kury.

Last week, her hospital’s internal testing lab finally came online and is now turning around Covid-19 tests within 12 to 24 hours, says Kury. Before that, she was waiting up to a week for results from Quest Diagnostics, a large commercial lab which a recent investigation by The Atlantic alleged has contributed to California’s current testing backlog. (Representatives from Quest did not return a request for comment.) Since the hospital’s own lab started analyzing results, at the two Alta Bates campuses in Oakland, health care workers went from using 6,000 N95 masks per day to 1,000, according to Kury. “Turnaround time made a huge difference in being able to spare what is becoming increasingly sparse PPE,” she says.