Solutions (Quality and Warranty)

Quality and Warranty

Around the world, leading companies use PolyVista's solution to identify emerging product warranty and quality issues.

While there is no such thing as a typical PolyVista user, most users involved with warranty/quality are domain experts that often have deep knowledge of their company’s products and manufacturing processes. They are often experts, may have Six Sigma training, and are – by nature – curious and dogged in their efforts to track down a problem.

Before PolyVista, these users were often frustrated by the demand for quick answers, the ever-increasing amount of data coming at them, and the need to help set priorities for the limited resources available. In many cases, they faced a skeptical audience that either said or thought, “Prove it,” when they tried to raise an alert about an emerging problem.

HP’s Brian Walker, Director of Automotive Industries for Worldwide Manufacturing and Distribution Industries, stated, "[PolyVista] reduced the time it takes to detect problems in many product areas from 120 days to 24 hours."

The strength of PolyVista's technology is that it delivers a virtual speed-of-thought analytic experience. PolyVista seamlessly integrates analytics, visualizations, and data and text mining in one graphical interface. The user interface provides an environment where users can interact with their data to:

“It isn't that they can't see the solution. It is that they can't see the problem.”*

G.K. Chesterton
Early 20th century English writer, poet, philosopher

  1. Confirm assumptions or prior knowledge
  2. Answer specific questions, and
  3. Use prebuilt algorithms to automatically mine their data to uncover emerging issues earlier than previously possible.

Results can then be explored to confirm new findings, and the analysis can be shared using integrated interactive visualizations to graphically illustrate and inform others about emerging issues and results of corrective actions.

As the director of engineering at Trane Corporation stated:

“Our quality engineers built their careers on their pivot tables. We were concerned that they wouldn’t take advantage of the capabilities PolyVista has to offer. We were pleasantly surprised that the quality engineers at each of our 23 factories quickly saw the advantages of using PolyVista. With PolyVista, they were able to quickly answer a question or uncover a quality issue and get on with the job of resolving those issues rather than searching for the problems. Instead of spending time trying to find a problem, they were able to spend their time fixing problems.”

Typically, data comes into PolyVista in multiple forms and from multiple sources. The data can be structured (numbers) and unstructured (free-form text). The data is organized in a hierarchical manner. As an example, geographic data could be organized by continent, country, state, county, and city, down to individual street addresses. The same data could also be organized by markets, distribution channels, retail versus wholesale, or any other organization that is useful to the user.

To use PolyVista, you don’t need to predefine rules or prebuild models. After all, how can you define or build a model about something that has never happened before? The software automatically scours your data, including free-form text, looking for undefined anomalies, relationships, correlations, and patterns and then presents the results for the user to review.

To ask a question of the data, a user simply has to drag and drop a data element into the Excel-type workspace. Users can then explore, drill into their data, add additional data elements to see the data from different perspectives, slice and dice the data, and/or graph the results for presentation – all on a single screen. On the same screen, with only a few clicks, a user can launch an algorithm that will automatically explore the data and return multiple results.

The prebuilt algorithms in PolyVista are easy to use and require no statistical training or building of rules. Even though PolyVista users are not statisticians, a powerful statistical engine operates in the background. Algorithms like Statistical Process Control (SPC) to perform automated trending analysis, Discovery (Chi-square) to

"PolyVista’s software doesn’t get in the way of doing our analysis – it’s transparent to the user."* Tim Logan, Director of Flight Operational Safety Southwest Airlines

Tim Logan, Director of Flight Operational Safety
Southwest Airlines

automatically identify anomalies in data, Deviation to identify variances from standard or normal, Difference to automatically compare data from one period against another period to uncover an emerging issue, and Full-Text Search (FTS) to identify a simple or complex set of words or phrases in data are just some of the prebuilt algorithms that are all launched with a few clicks by simply selecting areas of interest based on a user’s knowledge of their business.

A user can then review the results and drag and drop the most interesting results into the workspace to further explore the information. Results can be exported to Microsoft Word, Excel, or PowerPoint or shared with another user by simply sending a link to the analysis. All these options are found in an easy-to-use interface that can usually be learned with about ten minutes of instruction. Why is so little time needed to become a PolyVista user? Because the user already knows their data! PolyVista’s single-screen interface is uncomplicated to use and built to facilitate analysis without getting in the way.

"We used the Discovery algorithm to find things that maybe didn’t make sense. That was kind of the approach we took – if the Discovery algorithm finds something that doesn’t make sense, maybe that’s where we should start digging to see why it triggered,” explained Brian Brown, Vice President of Customer Advocacy and Research for Citibank.

Southwest Airlines’ Director of Flight Operational Safety, Tim Logan, said, “PolyVista brings together both our structured (numeric) data and our unstructured (textual) data from our incident reports. It provides an easy-to-use environment that enables our staff to not only ask a question and get an answer but also automatically guides our users to new and unexpected information hidden in our data. The automated discovery function of PolyVista increases our analysis capabilities without having to add additional staff. It’s like adding really smart and experienced analysts.”

PolyVista’s technology has two basic functions: to confirm suspected conditions or situations and to discover unknown or unforeseen conditions and situations. PolyVista fills the gap between reporting and high-end analytics by enabling business users to discover new and valuable business insights hidden in their numerical (structured) and textual (unstructured) data.

Dr. Roberto states, "[Leaders] do not wait for problems to come to them.” They realize "every organization, no matter how successful, has plenty of problems, [and those problems] often lie beneath the surface, hidden from view. They behave much more proactively. They seek out problems. They embrace them...The very best leaders know that speed is critical. The earlier you discover a problem, the more likely you can contain the damage, and the more likely you can solve it readily. Most important of all, successful leaders do not see problems as threats. They see every problem as an opportunity to learn and improve."

PolyVista helps leaders discover opportunities to learn and improve.

Contact Us