Haller Information Technology System (HITS™)

HITS™ includes the complete package of tools required to achieve robust process improvement results in one easy to use Windows™ compatible software system. HITS™ uses the proven Haller 5-Step process for identifying and controlling the important variables, conducting efficient statistically designed experiments, analyzing the results and optimizing the system performance to achieve the best balance of properties. A 3 day training program is offered to develop and enhance skills in product and process improvement using the proven HITS™ strategy.

HITS™ introduces a host of new powerful features such as:

  • the ability to benefit from experiments and plant trials conducted in the past.
  • the provision to easily exclude undesirable experimental conditions from the plan.
  • the ease of determining complex models of processes which enable users to predict future outcomes accurately.
  • the means to efficiently gather the data required to identify the principal sources of variation.
  • the use of a "Goodness Equation" to optimize many properties simultaneously and drive each to its desired value.
HITS™ was based on years of hands-on experience working with all levels of engineers and scientists on the "plant floor." The single focus was to achieve new levels of product performance and manufacturing productivity. Numerous companies have used the HITS™ modules to:
  • Significantly reduce product invention cycle time
  • Markedly improve product properties
  • Dramatically increase manufacturing productivity.

Process Optimization Strategy

While each of the five HITS™, Windows™ compatible software packages can be used as a stand-alone tool, more problem-solving power can be achieved by using them in conjunction with each other as shared applications.

Control Chart Analysis/ Analysis of Means (HITS™-CCA/ANOM)
Measures how much variation is occurring in a process when the data are time ordered. Is the variation occurring short-term or are significant shifts and trends occurring due to time related variables? Is there a significant difference between the performance of 2 or more groups when measured by the average?

Sources of Variation Analysis (HITS™-SVA)
Quantifies the variation due to various sources in a process. Using balanced or unbalanced nested data gathering plans for 2 or more components, SVA© computes the variation associated with each component. This enables process improvement efforts to be prioritized.

Experimental Design Optimizer (HITS™-EDO)
Constructs experimental plans that permit data on many variables to be gathered in a very efficient yet highly informative and effective manner. EDO© designs are generated based on proven D-optimal methods that converge on designs from which models can be developed with prediction accuracy equal to experimental error. EDO© designs are not selected from a library of existing designs. On the contrary, each EDO© design is generated based on the user's definition of the project. Design spaces can be constrained to avoid including low information experiments in the plan. Existing data can be used as a basis for generating the D-optimal experimental plans in order to minimize the number of new experiments required and accelerate development efforts. Designs can be generated for any model as specified in terms of variables (non-mixture and mixture) and their interactions.

Multiple Correlation Analysis (HITS™-MCA) Analyzes historical data or data from experimental plans in order to develop process models. In turn, these models can be used to improve and control processes for optimum performance. Models can be based on known theories or empirically developed. Analysts using MCA© can identify outliers and leverage points in the data as well as regions for further experimental investigation. Predictions from models can be generated in MCA© and then simply cut and pasted into spreadsheet programs in order to present the analytical results graphically.

Multiple Property Optimization (HITS™-MPO)
Finds the best balance of process settings and raw material conditions to optimize quality, productivity and cost simultaneously. Analysts can easily include subjective definitions of property values into the software. MPO© uses models generated in MCA© or business models to identify process conditions for optimum performance consistent with the customer's definitions of value. Grid search techniques for locating process optimums avoid the problems associated with self-directed search methods that converge on local optimums rather than gobal optimums.

Which HITS™ Tools Should Be Used?
Determines which processes are varying beyond acceptable limits and when this occurs. Determines which process averages are significantly different.
SVA©; - Determines and quantifies the sources of variation in any process.
EDO©; - Develops a plan for collecting data about any process and its raw materials in an efficient manner.
MCA©; - Extracts and quantifies all meaningful correlations based on data mined from historical records or gathered according to an experimental plan.
MPO©; - Optimizes all properties simultaneously particularly those that conflict with each other.

Harold S Haller & Company
31004 Nantucket Row, Cleveland, OH 44140-1050
Phone: (440) 871-6597; Fax: (440) 871-1889
E-mail: Hshaller@aol.com