Developing better products
faster
|
INForm - Modeling and Optimizing Formulations
INForm enables product
formulators to optimize their formulations to meet specific goals, saving time
and balancing costs against product performance. With INForm you can
tailor your formulations to each particular customer or project - and you can
determine quickly when you've reached the best solution. No more time wasted
exploring unfruitful ideas, and in trying to improve formulations that can't be
bettered within your existing constraints and product quality
requirements!
"I find INForm to be one of the most versatile programs available for advanced data analysis. The program allows me to systematically explore a design space whereby I can collect new data and feed it into the ever-growing model. I can take steps in new directions with my experiments and INForm is able to help me map the overall design space so that I have a robust understanding of my new product formulations."
- Michael R, PhD, Senior Research Scientist, USA
Because INForm is 'data driven'
by the experimental results you feed into it, it can be used for a range of
challenging problems in pharmaceuticals, health and personal care, home care,
paints and coatings, adhesives... wherever product formulation has an important
role to play. And, in each of these areas, it can tackle a variety of problems,
making it an indispensible part of the formulator's tool-kit that will be used
again and again.
INForm can work over a LAN, so
can facilitate information-sharing with your colleagues, improving the
teamworking necessary to solve complex problems. It's an essential component of
any knowledge management system for formulation.
How INForm works
INForm uses neural network
technology and Gene Expression Programming to develop models, discovering the relationships buried in your
experimental data. These techniques can capture subtle and non-linear
relationships, without requiring you to pre-judge what form the relationships
will take. Multi-variable formulations are easily handled, so you do not need
to risk losing important information by oversimplifying your problem.
Once you have your model, integrated
statistics and graphics (both 2D and 3D) give you intuitive guides of how to
improve your formulations. And you can try out 'what if' ideas to see what will
work best, and how changing ingredients and process conditions affects
properties.

But, to get the best formulations
quickly, use INForm's optimization capability. Based on a combination of
genetic algorithms with fuzzy logic, the optimizer is intuitive and easy to
use. You set up an 'objective function' based on the values you want for each
property, and on the relative importance of the properties. You can even define
your own properties - like the cost of the formulation - that are mathematical
functions of the inputs. Once you have specified your objectives, the genetic
algorithm searches the multi-dimensional space quickly, assessing the 'fitness'
of each solution and developing new solutions to find the best.
Finds models automatically from your data
INForm is 'data driven' - it
uses your own experimental data. As long as you have data, collected in a
spreadsheet-type format, you can import it directly into INForm. Simply
specify which variables are inputs (ingredients/processing conditions) and
which are outputs, so that INForm can develop its cause-and-effect
models for you. INForm does not require you to gather specific new data.
You can use the historical data already collected within your organization, or
start with a small study to scope out the problem. It is a simple matter to add
new data later, and quickly develop better models.
ANOVA statistics, integrated into INForm, let you assess the quality of your models. As with all modelling
techniques, the better the data you provide, the more reliable the answer you
will get.
Optional Support & Maintenance is provided with INForm. We also offer one-day training courses on using INForm.
To access the Application Notes showing
how INForm can be used, you can follow this
link.
|