Avidbots shares data insights with Management Analytics students at Lazaridis School

March 21, 2023

At Avidbots, a leading global robotics firm based in Kitchener, Ont., the quality-assurance process is data-driven — and it closely resembles the work students in Lazaridis School of Business and Economics at Wilfrid Laurier University’s Master of Science in Management Analytics (MMA) program may soon engage in after graduating.

To hear more about how Avidbots uses data analytics to improve its products and processes, the Lazaridis MMA program hosted a seminar on March 2 with Dennis Doroslovac, the company’s director of quality and information.

Doroslovac is also a Lazaridis school alumnus (MSc. ’21), graduating from the Management of Innovation and Technology program. His presentation to MMA students was a deep dive into the practical impact data science has on business outcomes.

Here’s an overview of the insight he shared.

What is quality?

Ask five people, and you’ll likely receive five different answers.

But Doroslovac defines it as: the degree to which a product or service meets the expectations of a customer.

“If you're satisfied as a customer, we know we've been able to achieve good quality,” he said.

“If you're dissatisfied, something happened along the way ... that was disruptive to you.”

Why pursue quality?

Ultimately, it generates profits by capturing unnecessary costs.

In addition to external stakeholders — people who buy an autonomous Avidbots floor-cleaning robot — the company has many internal customers.

These includes the production department, the finance team, the service team, and others. It’s important to use data analytics within all these company segments to reduce costs.

“We want to make sure that everything we build in that robot functions and performs exactly the way it's designed to,” said Doroslovac. “And quality has a major role to ensure that end line is met.”

A data-driven approach to quality

Quality, speed and cost are interrelated.

A successful quality program leads to increased productivity and reduced costs. It also ensures mistakes in the manufacturing process aren’t repeated.

Poor quality anywhere in the chain creates repeated mistakes, which leads to wasted time and resources to fix them.

“Eventually, [quality is] going to increase sales, market penetration and profitability,” said Doroslovac. “I firmly believe they're all tied together.”

The cost of quality

Data analytics can quantify waste — everything from unnecessary labour due to a faulty manufacturing process, to product failures in the field, warranty claims and work orders.

At Avidbots, a work order usually indicates the company was required to deploy a technician to a customer site to carry out repairs.

“It's a massive cost on our end every single time we have to address a problem in the field,” said Doroslovac.

“If you've had good quality, and a robot that functions properly, you're going to minimize, reduce, or eliminate most of these causes to occur in the field.”

Returned goods are barometers for quality

This refers to components that are returned to the manufacturer.

Avidbots relies on external suppliers for some of its components, and if one is returned, it sparks a company-wide review to determine if its other robots in the field have the same issue.

“There's a massive cost connected to that.”

Data selection connects to business goals

Analytics are powerful, but only if a company is collecting the right data.

Relevant data is that which connects directly to a company’s business goals and strategies.

At Avidbots, staff uses the SMART framework — an acronym for Specific, Measurable, Achievable, Relevant and Time-Bound — to guide its data selection process.

“You can collect data for anything,” noted Doroslovac. “And a lot of it [various types of data] doesn't really matter … or it matters so low, versus the effort and resources that you apply to something that could attain a bigger win for the organization.”

A bird’s eye view

One of Doroslovac’s strategies is to take a high-level overview of the entire manufacturing process, noting the total quantities of various metrics registered in a given month or year.

Then he and his team dive deeper and pinpoint specific factors leading to negative outcomes.

He believes firmly in the Pareto Principle, also known as the 80/20 rule — the idea that 80 per cent of the results come from 20 per cent of the effort.

“I cannot stress to you how important it is — Paretoing our graphs within our organization,” he said. “It helps us align and prioritize what's the most significant issue we need to attack?”

Data-driven leadership

Collecting and analyzing data is half the battle; the rest is making correct decisions at the management level.

In his 20-plus years in for-profit manufacturing, Doroslovac has found that money talks.

“If you've got something that's performing poorly and it costs money, and it's against the strategy around increasing revenue, people will listen,” he said.

“You will gain a lot of attention from your peers, from leadership, from the owners, from everybody, if you're speaking in terms of the financial success.”

Problem solving is for everyone

Not everyone is a data scientist, but all employees are problem solvers.

“You have to empower everybody in the organization,” said Doroslovac, noting employees also need enough time to bring about a satisfactory result.

“A good problem-solving event, for me, it takes almost a month to do. But after that, the expectations of that is, the biggest problem that you had, it will never come back again. It's eliminated completely.”

Providing a bridge between theory and practice

This seminar is part of a series aimed at helping MMA students connect with leaders in various industries who use data in constructive ways to improve products and service delivery.

“Often, when people think about data analytics, data science and AI, they're sort of hooked on the big tech companies, or marketing and finance,” said Michael J. Pavlin, associate professor, director of the MMA program and William Birchall Chair in Management Analytics.

“We're trying to expose [students] to this big range of data analytic capabilities beyond that.”

The Lazaridis School thanks Dennis Doroslovac for the time he gave and  the lessons he shared with our MMA class. Learn more about the Masters of Science in Management Analytics program.