Patented Innovations in Design of Experiments
MoreSteam's patented innovations simplify the process of designing, evaluating, and analyzing experiments without compromising statistical rigor.
MoreSteam Innovation Yields Multiple Patents for Design of Experiments
For more than two decades, MoreSteam has developed technology designed to make advanced statistical methods more practical and accessible for people solving real-world problems. In 2011, MoreSteam was granted a US patent for inventing a software automation to assist with hypothesis testing - so-called “hypothesis testing in plain English,” an approach that was adopted by many others. Innovation in user-centric design continued, and our work in Design of Experiments (DOE) has resulted in multiple U.S. patents covering innovations in experiment design, statistical analysis, and the way practitioners interact with DOE software.
These patents reflect the same philosophy that guides the development of MoreSteam software today: put sound statistical methodology behind the technology so practitioners can spend more time understanding their processes and making better decisions.
Taking the guesswork out of designing an experiment
U.S. Patent 8612366 — Systems and Methods for Performing Design of Experiments
Setting up the right designed experiment requires a series of important decisions. How many factors should you test? How many runs do you need? Should you use a full or fractional factorial design? What about replication, blocking, or multiple factor levels?
MoreSteam’s patented approach guides practitioners through those decisions step by step. The software asks questions about the experiment and uses established DOE conventions to recommend an appropriate design, helping users account for factors, levels, runs, blocking, replication, responses, and statistical power.
Instead of requiring practitioners to know which experimental design to choose before they begin, the system puts that expertise into the software. The result is a more approachable way to build statistically sound experiments while helping users understand the decisions behind the design.
Inventors: William Hathaway, Smita Skrivanek
Filing date: 09-29-2010
Issue date: 12-17-2013
Know what your experiment can detect before you run it
U.S. Patent 9965721 — Systems and Methods for Designing Experiments Based on Statistical Power
An experiment can be perfectly executed and still fail to answer the question it was designed to answer. If the design doesn’t have enough statistical power, an important effect can be lost in the natural variation of the process. Unfortunately, this is something practitioners may not discover until after they’ve invested the time and resources to run the experiment.
MoreSteam’s patented approach brings statistical power into the experiment design process. The software evaluates different combinations of effect size, replicates, and experimental runs, then shows how those choices affect the experiment’s ability to detect meaningful changes. Practitioners can compare design options, including the potential cost of additional runs, before deciding how to proceed.
Instead of finding out afterward that an experiment wasn’t capable of detecting the effect that mattered, practitioners can make a more informed choice from the start, balancing statistical confidence with the practical cost of experimentation.
Inventors: Lars Maaseidvaag, Smita Skrivanek, Alan Gao
Filing date: 09-16-2014
Issue date: 05-08-2018
Making Design of Experiments easier to interpret
U.S. Patent 9007375 — Systems and Methods for Presenting Results of Experiments
Design of Experiments can uncover which factors or combinations of factors have the greatest impact on a process. However, running the experiment is only part of the challenge. Practitioners also need to make sense of the results and determine what deserves their attention in the statistical output.
MoreSteam’s patented approach automatically analyzes experimental data to identify significant factors and interactions, build a recommended statistical model, and present the results visually. The method preserves model hierarchy while clearly distinguishing significant effects from those that are not significant, helping practitioners move more quickly from statistical output to practical understanding.
The result is a more guided approach to DOE analysis that helps users focus on what the experiment is telling them about the process, rather than leaving them to sort through statistical output on their own.
Inventors: William Hathaway, Smita Skrivanek
Filing date: 09-29-2010
Issue date: 04-14-2015
Using MoreSteam’s patented methods
MoreSteam makes these patented methods available under an open license because we believe good statistical methods should help more people solve problems effectively.
If you provide goods or services that use or incorporate the systems and methods covered by these patents, you are required to acknowledge MoreSteam as the source and include a link to MoreSteam.com. This attribution helps others find the original work and learn more about the thinking and technology behind it.
Suggested attribution:“Certain systems and methods used herein are based on patented technology developed by MoreSteam.com LLC or a MoreSteam.com LLC affiliate.” Link “MoreSteam.com LLC” to MoreSteam.com.
For questions about the patents, licensing, or appropriate attribution, please contact us at DOE@moresteam.com.