Published June 16, 2017
Meals on Wheels for WNY produces and delivers between 5,000 and 6,000 meals a day — including hot and cold meals for a range of medically appropriate diets.
The meals are delivered through a multi-stage process to dozens of sites and, ultimately, 1,800 homes. Home-delivery clients receive a hot meal, cold meal, friendly conversation and well-being check daily.
“Our top priority is delivering a healthy, safe, medically appropriate meal to each client every day — our logistics, operations and workflows are absolutely crucial,” says Meals on Wheels for WNY Chief Operating Officer Chris Procknal.
So when Meals on Wheels was contacted in January by three UB industrial engineering graduate students who wanted to talk about including the organization in a project centered on advanced analytics and organization science, Procknal and other Meals on Wheels executives agreed.
“We called our project ‘Pro Bono Analytics,’” says Vineet Payyappalli, a fourth-year industrial engineering PhD student specializing in operations research. “We are in the student chapter of UB INFORMS — the Institute for Operations Research and Management Science.”
Nationally, INFORMS has a division for pro bono analytics, so the UB graduate students started a plan to do their own project here.
“I went online and went to Yelp, looking at different non-profits, how they were rated and what people had to say,” says Jessica Dorismond, who is also in the fourth year of a PhD program in industrial engineering.
“For Meals on Wheels, deliveries are key to what they do,” she explains.
Payyappalli and Dorismond were joined by Prabakar Theivaraaj on the project’s first phase during this past spring semester. Theivaraaj holds an MS from UB in industrial engineering.
The trio met with members of the Meals on Wheels leadership to learn details about their delivery process, their challenges and to see if there was a way they could help.
The team decided to focus on optimization of the pack-out process for delivering meals. Every Meals on Wheels home-delivery route changes daily, so the counts and daily diet types are in constant flux. There is a strict two-hour safe window between final plating in the commissary and delivery at the last client’s home.
“We have five diets — regular, renal, ground, bland and diabetic — and when each new client comes on we work with their doctor to determine what diet they should have,” Procknal says. “As you can imagine, incorrect deliveries of any meals can have a serious consequence for our clients.”
After a hot meal is sealed at the end of the assembly line, it is checked against a list — sorting the meals against different routes and sites by the recipients’ addresses, ensuring that meals go into the correct ovens for delivery.
Payyappalli’s team found the checklist did not have the different types of meals supplied to clients. The list did have the total quantity for each route and for each car for delivery — but did not have the sub-quantities of the meal categories.
Theivaraaj says analytics conducted over the mid-portion of the semester enabled the UB team to create new lists for consideration by Meals on Wheels for WNY leadership.
“We set them up by order of delivery — as the meals came off the assembly line — as opposed to alphabetically listing the neighborhoods and communities,” Theivaraaj says.
Procknal notes that after a two-week trial with the new pack-out template,
“the analytics are a very good fit with our operations systems,” allowing for more accurate tracking that will enable staff and volunteers to serve clients more efficiently.
“The food and friendship we bring to these homebound individuals is absolutely crucial in ensuring that they can live with dignity and independence,” he says.
“Our thanks go out to this group of UB engineers for their passion and dedication in helping us optimize our meal pack-out process.”
Ann Bisantz, professor and chair of UB Industrial and Systems Engineering, says that in the department, “We directly connect research with learning and societal impacts. We support student-led, hands-on learning activities,” she says. “Our Experiential Learning Programs help students develop professionalism and a practical perspective by connecting the classroom to the real world.”
Dorismond says she also found out about The Teacher’s Desk by searching Yelp. “We immediately connected with their mission to distribute free school supplies to students in need,” she says.
Payyappalli, Dorismond and their team leader, Yashar Khayati, who had just received a PhD in industrial engineering, went to The Teacher’s Desk, a small not-for-profit in Buffalo, last January to introduce themselves and talk about “Pro Bono Analytics” with the owner, John Mika.
Khayati says they discovered that Mika’s biggest challenge — organizing The Teacher’s Desk’s huge, constantly changing inventory — fit their project perfectly.
This year Mika and The Teacher’s Desk distributed more than $6 million in free school supplies to more than 100,000 students in 230 schools across nine Western New York counties.
“It’s set up like an office supply store in a donated warehouse,” Mika says, “and allows teachers from area schools to shop for free supplies, all of which are also donated.
“We have 15,000 square feet of warehouse space. The big challenge has turned out to be that because this huge inventory changes week to week, none of our 200 weekly volunteers are able to keep up with what is stored where,” he says.
The Teacher’s Desk stocks items ranging from pencils, copy paper and crayons to backpacks and picnic totes for field trips. Items are new or gently used, manufacturing by-products or overstock. “We needed a system that would work to organize a constantly changing inventory,” Mika says. “I’m usually the only one who knows where to find stuff, along with my two full-time staff members.”
Khayati says Mika needed a layout design that was flexible enough to take that into account, and that would work with his stock at any given time.
Together with Tianjiao Wang, a second-year industrial engineering PhD student, Khayati measured Mika’s warehouse space. The pair created a program that named each area of the store, identified the space and recorded the contents of each shelf.
Praveen Frederick Selvaraj, who holds an MS from UB in industrial engineering, and Sarathkumar Nachiappan Nallusamy, who is in the first year of an MS degree program, also were part of the team.
“Yashar and our team created a mathematical program where, say, a shipment of pencils comes in, and you enter the total weight and the number of pencils,” Wang explains. “The program then shows available spaces and which ones would work for that item.”
“If you are looking for pencils, you look them up, check the number and go there to pick up what you need,” Selvaraj says. “As new items come in week to week,” Nallusamy adds, “they are assigned to spaces that held something else the previous week, so everything is always easy to find.”
Mika calls the UB engineering team’s work “impressive.”
“We find a use for everything, but UB found a place for everything,” he says.
“It is inspiring to see young people who are so passionate about what they are doing. Their efforts will allow us to be more efficient when we start up again for the fall semester.”