This surface plot shows the difference between conventional models of consumer perceptions and the more complex (and accurate) models that can be developed in JMP.
Washing away R&D worries in the consumer products industry
DETERTEC uses JMP® to virtualize consumers’ visual perception, helping clients save time and money
Now more than ever, business competition is fierce. The overarching goal dictated by executives is the need for steady top-line growth, creating the need for R&D teams to continually accelerate their pace of innovation to stay ahead.
In the consumer products industry, the story is no different. Companies are compelled to design products and technologies that offer a superior perceived consumer value. The most successful products in this highly competitive market offer improved sensory experiences to demanding consumers.
However, these products not only need to be better in consumers’ eyes but also cheaper to produce. How do companies tackle this seemingly impossible challenge? By partnering with companies like DETERTEC, a small entrepreneurial technology company that specializes in the areas of detergency (the quality of cleansing), spectroscopy and vision science.
Established in Quito, Ecuador, in 1994, DETERTEC acts as a service provider of R&D activities for many regional and global consumer product companies, focusing on laundry markets. With clients such as Procter & Gamble (US), Unilever (UK), Henkel (Germany), and CIBA (Switzerland), DETERTEC has become the regional leader in providing highly specialized services for leading regional and global soap and detergent companies.
Since many traditional R&D efforts used by these companies can often take two or three years to complete, the amount of money invested quickly adds up. And at the end of the day, if the final product does not offer benefits from the consumer’s point of view, then the money was not well spent and did not result in a cheaper product.
“With time and cost savings in mind, it is expected that as much as 90 percent of R&D for consumer products will be done in the virtual world in coming years,” says DETERTEC Director General and co-founder Rodrigo Olmedo. “Thanks to the application of new computational models, the use of ’in silico’ research – or research performed via computer – is quickly surpassing conventional in vivo and in vitro R&D practices.”
Challenges of the past
But this wasn’t always the case. In the past, DETERTEC encountered several challenges in the area of detergency research:
Conventional spectrometric techniques used to characterize fabrics had limitations for providing relevant consumer data, with no flexibility for modifying hardware and software components.
Existing mathematical models were outdated and did not adequately correlate with perceptions of 21st century consumers.
Conventional in vivo and in vitro methods were expensive, putting pressure on R&D to find more cost-effective alternatives.
According to Olmedo, one common theme arising from these frustrations was the need for a software solution that could manage large amounts of data, with scripting and modeling capabilities, extensive three-dimensional graphical features, and the ability to program a large sequence of events.
Cleaning up with JMP®
The answer: JMP statistical software from SAS. A JMP user since 1996, Olmedo did not consider any other software for DETERTEC’s needs. Fortunately, says Olmedo, “It was a happy coincidence that JMP version 4 included scripting capabilities and JMP version 5.1 featured interactive three-dimensional capabilities - precisely when we needed them.”
According to Olmedo, JMP has a user-friendly interface and a great environment for graphing data, and evidently others around the world agree. “Since our developments are a collection of scripts developed with JMP, when our clients implement our spectroscopy or computer-based technology, it’s mandatory that they have JMP. Many of our clients already have it, and if not, it’s easy for them to implement. At DETERTEC, we are helping to promote JMP worldwide.”
The LaundryVisualLab™ solution
In 2005, DETERTEC launched a new virtual visual consumer experience, an innovative modeling and simulation system for the laundry industry called LaundryVisualLab™. The system is based on the fact that consumers’ perception of detergent performance depends on their perception of the fabric’s visual properties after the washing process.
More specifically, explains Olmedo, product performance depends on consumers’ perception of cleanliness and whiteness in relevant laundry scenarios. But not only that -- in some specific laundry markets, customer value is also linked to consumer perception of wet fabrics and washing liquids at the beginning and end of intermediate washing stages.
The perceived value of laundry products, from a consumer perspective, depends on a matrix of several elements: consumer visual abilities, consumer preferences, sources of illumination with different relative UV content, other components of the environment, and different visual properties of luminescent and nonluminescent fabrics, among others. All of these elements need to be part of the virtual visual consumer washing experience along with the potential to virtualize product or technology benefits.
The LaundryVisualLab system is a collection of JMP-based scripts that use extensive data table manipulation, modeling, interactive graphs and visualization to apply radiation transfer, fluorescence emission, color appearance and consumer perception models. The final system is fully customizable to different inputs and outputs according to the technical information available in each case and preferences for visualization of results.
Module 1: fabric virtualization
The system contains two modules. Module 1 virtualizes fabrics before, during and after the washing process. A fabric has specific visual attributes relative to surface reflection and luminescence properties, which depend on the radiation energy in a specific visual scenario. Unfortunately, conventional equipment to characterize surface properties offers limited information from the point of view of real scenes and actual market conditions.
DETERTEC overcame this limitation with the use of recent developments in photonics and thin film technologies to measure absolute radiance coming from a fabric under actual sources of illumination. This innovative type of measurement provides a massive amount of data.
“We selected JMP to process that information because of its powerful scripting capability to manage tables, as well as its linear and nonlinear regression models and its ability to do complex calculations,” says Olmedo.
Module 2: consumer cleanliness and whiteness perception
The second module virtualizes consumer cleanliness and whiteness perception under actual visual environments. This is a critical parameter for building a virtual environment that includes visual perceptions. “It is well known that mathematical models developed in the 1970s have some level of correlation with panelists’ visual perception,” Olmedo explains. “But its lack of correlation with real consumer perception in actual market scenarios is also well known, and the problem is bigger if the intention is to model and simulate the global laundry market characterized by several different regional markets, with different consumer preferences and extremely different visual scenarios.”
For example, says Olmedo, the perception of fabric whiteness in Asia is different in comparison with that in Latin America because of changes in the visual environment (such as lighting) and differences in the fabric itself. “White fabric in Asia is not the same as white fabric in Latin America.”
DETERTEC’s approach was to develop a flexible modeling procedure to represent consumer perceptions by using analytical and geometric modeling systems to develop several customized stain removal indexes that represent consumer cleanliness perceptions, as well as several customized whiteness indexes that represent consumer whiteness perception.
“This problem is very complex when working with very large nonlinear models and numerous parameters,” says Olmedo. “Thanks to recent developments in JMP’s capabilities for interactive plotting of several three-dimensional surfaces, it is possible to find appropriate parameters that shape quadric response functions like tilted two-sheets elliptic hyperboloids and tilted ellipsoids to reach levels of correlation of more than 95 percent with actual consumer responses.”
The results: “a very competitive advantage”
According to Olmedo, the LaundryVisualLab system has the potential to become the link that closes the chain between the physical properties of fabrics and consumers’ positive visual experiences in different locations around the world - and ultimately customer values, which are the purpose and focus of the consumer products industry.
With JMP’s capabilities, Olmedo says DETERTEC can develop complex models and perform calculations at tremendously faster speeds, all for a very affordable price.
“More importantly, now it is possible for DETERTEC to do things that were just not possible before, and JMP has been the key to our innovation process,” Olmedo says. “We have developed innovative spectroscopy and computer-based technologies to perform virtual R&D that is not only perfectly in line with what the industry needs, but also with what the industry is urged to implement in order to stay competitive.”
“We now have a very competitive advantage to offer our clients,” says Olmedo. “In order to stay competitive, our clients must continually implement new procedures, and we are offering them proven alternatives to modify their R&D time, cost and efficiency.”
“Our success can be quantified by the fact that we now have a more robust business position with our clients,” adds Olmedo. “We have reinforced our position as an innovation partner, and that is critical to furthering our business relationships.”
Now it is possible for DETERTEC to do things that were just not possible before, and JMP has been the key to our innovation process.
Director General and Co-Founder