'Our Hotel Industry Has Been Decimated,' Says VisitPITTSBURGH CEO

On today's program: VisitPITTSBURGH’s Jerad Bachar says it’ll take a mass vaccination campaign before the hospitality industry recovers from the pandemic; researchers are seeing how machine learning can help Pittsburgh Water and Sewer Authority find lead lines; and in light of studies showing in-person learning can happen safely, some say teacher unions are prioritizing teachers over students.

Hospitality industry will get a boost from state funding, but the region needs more

(0:00 — 6:34) 

Gov. Tom Wolf announced last week $145 million will be distributed across the state to the hospitality industry, one of the industries hardest hit by the pandemic and the recession.

A report from the Allegheny Conference on Community Development found the region’s own hospitality industry is in desperate need of a boost

 “Our hotel industry has been decimated over the last year, and it’ll soon be over a year,” says Jerad Bachar, CEO of VisitPITTSBURGH, the official tourism agency for the region.

Bachar says, echoing President Joe Biden, that this money is “a down payment.” 

“$145 million doesn’t go very far when you start divvying it up,” says Bachar. “It’s a wonderful program, it’s certainly gonna be welcomed by the community, but it’s not going to fix the problem.”

The problem, he says, is people need to have confidence they will be safe while traveling. Bachar says this can be achieved with a robust vaccination campaign.

Bachar says some local hotels recouped some lost funds by serving as dorms for university students. Others have tried to bring in other revenue by offering weekend packages to locals, like a “staycation,” or offering a day rate so those working from home can get a change of scenery. 

But these efforts don’t balance the fact that hotel occupancy levels have plummeted since last March, leaving millions of rooms empty throughout the year, affecting tax revenue in addition to hotel profits. 

“It really doesn’t impact a branded hotel versus an independent hotel more or less, it really depends on that particular hotel,” says Bachar. Every hotel operates and hires locally, says Bachar, so even if the hotel holds a Marriott title, the lack of business affects Pittsburgh residents. 

Despite the downturn, Bachar says VisitPITTSBURGH is still expecting some events this year, like the NCAA Frozen Four hockey event coming to town in April. Bachar says future events depend on the level of restrictions in the state and confidence in safety from visitors. 

“Live events are so important to us as a society, they’re so important to us as an economy,” says Bachar. “We can get our live events back up and running, but we need to do it sensibly.”

PWSA and researchers are using machine learning to try and predict the placement of lead lines
(6:37 — 13:25)

Since June 2016, the Pittsburgh Water and Sewer Authority (PWSA) has replaced more than 8,200 lead service lines, but both Mayor Peduto and PWSA officials have said on The Confluence the maps of all lead lines are incomplete, making the work more difficult.

Over the past year, PWSA collaborated with the University of Pittsburgh on a machine learning model to find and replace the lead lines.

“In our modeling work, we attempted to predict the locations of all the lead lines in PWSA’s service area, and we found the data and models weren’t sufficient to predict the entire inventory,” says Dr. Michael Blackhurst, a research scientist at Pitt’s Center for Social and Urban Research. Blackhurst is an engineer with expertise in water resource planning.

Instead of focusing on creating a complete map, Blackhurst and his team used machine learning and data to predict the likelihood a home is serviced by a lead line

Blackhurst says his research team used data such as when structures were built, historical records of materials, and tap water levels of lead in customer homes. They used these indicators to estimate the probability a customer’s service line is made of lead. The next, and more difficult step, says Blackhurst, is what probability you act on assuming a customer service line to be lead or non-lead.

“Is it 50%, is it 60%, is it 70%? And that involves a mix of judgment and exploration of the results.”

Blackhurst says the best application of the model is to plan a short-term cycle of replacements for customers who have the highest predicted probability of lead lines, which Blackhurst says is about 80% probability. 

Correction: In the interview, when our guest references 63% versus 73% when asked about the total number of lead lines in the city, they were refering to the accuracy of PWSA's historical data and the machine learning model, respectively.

Some say teacher unions are delaying in-person learning
(13:28 — 18:00)

More scientific evidence suggests that in-person learning, especially for younger students, presents a low risk for spreading the coronavirus. But last month, Pittsburgh Public Schools delayed reopening at least until April. 

90.5 WESA’s Lucy Perkins looks at the politics behind reopening schools.

The Confluence, where the news comes together, is 90.5 WESA’s daily news program. Tune in weekdays at 9 a.m. to hear newsmakers and innovators take an in-depth look at stories important to the Pittsburgh region. Find more episodes of The Confluence here or wherever you get your podcasts.