Increasing Profitability with Better Product Development Processes

“A lot of smart parts often add up to a dumb system,” says Dan Denlinger, founder of DED Systems and Master Coach for Six Sigma. With a Master’s Degree in Engineering Management from the University of Dayton and 20 years of systems engineering under his belt, Denlinger has the evidence he needs to form that conclusion. Now, with iThink an essential part of his toolkit, Denlinger leads clients through a whole-systems approach to development processes that satisfy customers and improve the bottom line.

DED Systems is named for Dan Denlinger’s initials. It also points to the practice’s value statement. “We bring dead systems back to life,” says Denlinger. “We’ve developed a process for improving product development, performing better testing faster, launching products effectively and on-time, and increasing profitability. We call it Design for Successful Systems. In the last 20 years, I’ve used iThink to apply that process to at least five dozen projects.”

Design for Successful Systems has evolved over time. Its original concept emanated out of a University of Dayton course that focused on reliability engineering. “We worked directly on product delivery projects,” says Denlinger. “The idea was to develop the ideal set-up for an engineering development system that minimized rework after launch.” (“Rework” is the effort that goes into warranty repairs or replacements and safety recalls.) The course also introduced him to iThink. DED Systems has partnered with Air Academy Associates to weave the Design for Six Sigma (DFSS) Process with the Design for Reliability (DFR) Process into an integrated lean product development process called Design for Successful Systems (DFSSys™).

I used iThink to design a better process that featured new policies for product development and testing

“After graduation I was hired by General Motors and worked in brake development,” says Denlinger. “Brake system recalls were costing about $30 million per year in 1990s dollars. I used iThink to design a better process that featured new policies for product development and testing. It took three years to implement the new policies but we wiped out brake system recalls. That immediately improved the bottom line.”

Since then, Denlinger has applied his process and iThink to many development process improvement projects. “It doesn’t focus on finance but the model is all about ‘show me the money,’” says Denlinger. “We’re always working to implement leaner systems that use the fewest resources to do the best job at product design, testing, and launch. The Design for Successful Development process includes 60 steps that coordinate product design, testing, and launch. That’s about 740 steps less than typical processes. One of our clients had a policy to staff 35 engineers on each product. Using our process, they were down to 5 engineers per product in five years. Their product launches were always on time, and they had eliminated product recalls. Those staff savings went right to the bottom line to boost profitability by 30%.”

Typically focused squarely on product development, his iThink model has also been applied to other operational processes that impact a business’ bottom line. One company used the model to design an internal IT help line. A lab that supports product development reduced their lead-time delay by 30% and an engineering department took its engineering staff from a fluctuation of 50 engineers over or under project requirements to 5 engineers over or under project requirements. “iThink allowed that client to look at hiring, training, and job rotation as a system in order to identify policy and process changes that led to that increase in efficiency,” says Denlinger.

Because [iThink] includes feedback loops, we can see consequences of system changes through time. Spreadsheets and other engineering tools take a fixed, or discrete, view of processes in time

The reason Denlinger’s model is able to address issues outside of product development is because it is process-based. And the benefit of taking a higher level process view is what keeps Denlinger modeling with iThink. “iThink is process-based,” says Denlinger. “Because it includes feedback loops, we can see consequences of system changes through time. Spreadsheets and other engineering tools take a fixed, or discrete, view of processes in time. We humans just don’t have the ability to anticipate consequences outside of our own experience. I’m reminded of that every time I use iThink because I’m always surprised. My mental models have never been completely right. iThink always provides insight into non-linear processes.”

The stakes are high when product development and other central business processes are revised. “Success testing is a plan to fail while failure testing is a plan for success,” says Denlinger. “iThink lets us test systems before we do bad things – launch a bad product or announce a new HR policy that doesn’t work.” Risks are further mitigated when process change decisions are put under the microscope of more detailed, discrete tools. “A development team can narrow its field of options quickly using iThink and then drill down on details with their fixed, point-in-time model,” says Denlinger.

Denlinger and his clients also appreciate iThink’s ability to affordably present a system visually. “Other solutions might offer a lot of animation features but they’re ten times the cost and impose a lot of processing overhead that can choke most computers,” says Denlinger. Import/export features allow Denlinger and clients to manage data and develop experiments in Microsoft Excel and then run models in iThink. And the dashboard allows Denlinger to turn control of models over to clients even if they aren’t trained in Systems Thinking. “Clients really like to drive the boat and see how the model is working.”

iThink lets us test systems before we do bad things – launch a bad product or announce a new HR policy that doesn’t work

By the end of an engagement, the iThink model has pointed Denlinger and his client toward process improvements that pay off in time, quality, and profits. “There’s no reason to be limited by a dumb system with smart parts,” says Denlinger. “By taking a whole system approach, we find the smart parts and make the system intelligent.”

STELLA Models Add Game-Changing Speed to River System Policy Planning

When using large amounts of data to model a large array of watershed management alternatives, speed is important. A STELLA model developed in the late 1990s as part of a shared-vision Apalachicola-Chattahoochee-Flint Basin Comprehensive Study is capable of running 70-year simulations in less than 5 minutes. That’s in contrast to the over 2 hours run time of the comparable, yet more detailed, simulations such as the Army Corps of Engineers ResSim model.

Steve Leitman, a researcher who has worked in the watershed for over 30 years and uses the STELLA model to investigate a number of water flow issues in the watershed, isn’t out to win a modeling time trial. Rather, he wants to give stakeholders including the Army Corps of Engineers, US Fish and Wildlife Service, NGO groups such as the RiverKeepers, farmers, and others the information they need to collaborate efficiently on management of a critical watershed.

Fed by the Chattahoochee and Flint Rivers, the Apalachicola River is a nearly 20,000 square mile watershed in the southeast U.S. that has greater average annual flow than the Colorado. The river system is an essential resource for three states: Florida, Georgia, and Alabama. Federal, state and local policies influence flow rates and water levels which in turn impact agricultural irrigation, navigation, and river biology. The complexity of the system, stakeholder disagreements over watershed management policies, and complicated data analyses impede decision-making.

“The STELLA model’s speed is game-changing,” says Leitman. “It’s much more efficient to discuss results when you still remember why you asked the questions. And fast models can help a team narrow down the number of options which should be tested with slower, more complex models.”

It’s much more efficient to discuss results when you still remember why you asked the questions

The shared-vision STELLA model was originally developed by Rick Palmer (now at the University of Massachusetts, then at the University of Washington) and his graduate students in cooperation with the states of Florida, Georgia, Alabama and the Army Corps of Engineers. The Army Corps of Engineers also developed a model in HEC-5 (now ResSim) to understand policy and management impacts on river flow and water basin levels. “We’ve kept the shared-vision STELLA model alive and calibrated it with the Corps’ model,” explains Leitman. “The two models allow us to take a collaborative approach to resolving contentious issues. The STELLA model delivers fast results that help refine assumptions and find options that address the concerns of all stakeholders. The Corps’ model offers a more detailed look at those options.”

Leitman points to three projects that have benefited from the speed and flexibility of the STELLA model in the last few years.

Reconciling River Navigation with Environmental Concerns

The Apalachicola is a federal navigation channel. When the State of Florida decided that it could no longer support dredging the river because of associated environmental impacts, environmentalists were happy. But there was a problem. Providing an authorized federal navigation channel is the major driver behind releases from the federal reservoirs. Without dredging, there was no commercial navigation, so the Corps stopped making the navigation releases. As a result, there were lower flows in the Apalachicola River, yet higher elevations in the federal storage reservoirs and in the Apalachicola River that meant a worse situation for both river ecology and boaters.

“Environmental interests were seeing navigation as the problem,” says Leitman. “The real problem was the way in which navigation was supported. There is more than one way to provide a navigation channel that is viable for commercial and recreational boaters. I started using the STELLA model so that we could look at the technical issues in the system and get away from the emotion. This project allowed the Apalachicola RiverKeepers and the Tri-River Waterway Development Association to look for modifications to rules for reservoir releases that would support navigation requirements and address environmental concerns.”

I started using the STELLA model so that we could look at the technical issues in the system and get away from the emotion

Finding a middle ground required “hundreds of simulations.” Given the speed of the model, that was practical and possible. Together, the stakeholders were able to identify and present the Corps with reservoir release options that eliminated or significantly reduced dredge maintenance needs, supported a seasonal navigation channel, and addressed environmental concerns associated with flow reductions.

Reconciling River Navigation with Environmental Concerns

Proactively Protecting Endangered Species

The US Fish and Wildlife Service (FWS) is responsible for tracking and protecting endangered species. When the Army Corps of Engineers makes a decision about river system management, the Fish and Wildlife Services reacts – typically to improve upon the plan from the perspective of fisheries protection.

“We thought, why not use the model to identify river management policies that would work for Fish and Wildlife rather than waiting for the Corps to propose their alternative and then having the FWS try to change it,” says Leitman. “Being proactive and providing the management agency with real alternatives is a better approach to advocating your management concerns than waiting to react to another group’s decision.”

Because the STELLA model runs so efficiently we were able to run selected alternatives in the morning and discuss the results in the afternoon

Again, the speed of the model was extremely beneficial. “Because the STELLA model runs so efficiently we were able to run selected alternatives in the morning and discuss the results in the afternoon,” says Leitman. “We also had the luxury of changing individual management parameters one at a time to develop a better understanding of how different reservoir control measures effect fish and wildlife resources among FWS staff. It only took 3 months for the FWS to develop and offer an alternative to the Corps for inclusion in their Water Control Manual updating process.”

Solving the Problem of Irrigation Water Withdrawals from the Flint River

Flow into the Apalachicola River is provided by the Chattahoochee and Flint basins and irrigation demands in the Flint basin have increased exponentially with the advent of center-pivot irrigation. Over the past 10 years researchers at the University of Florida and the University of Georgia have been field testing various approaches to substantially reduce irrigation withdrawals while maintaining the productivity and economic viability of the farms. Using the STELLA model, Leitman has been working with the University of Florida researchers to translate the effects of irrigation reductions on the entire basin

“We run the model with data from 70 hydrological years – 1939 to 2008,” says Leitman. “We’re also modeling a range of possible agricultural demand levels and then seeing the impact on Flint and Apalachicola River flow, reservoir elevations and performance metrics related to aquatic species including federally listed mussels and sturgeon and extent of floodplain inundation.” The group will continue the project by modeling groundwater impacts on river flow, climate change scenarios, and agricultural actions that can reduce irrigation demands.

The solutions we use [the STELLA model] to find are based on real data and a shared understanding of how the system will work

The STELLA model has shed light on difficult questions associated with managing a complex river basin. As impressive as the model’s speed is its continued utility in working through complex, often contentious issues is more impressive. “We’re often too focused on making increasingly complex models rather than using tools we have effectively and interpreting results,” says Leitman. “This model is very flexible and has been extended over time. It allows me to be a technical rather than an emotional advisor. The solutions we find using the model are based on real data and a shared understanding of how the system will work.”

Getting Kids off the Couch and Into Healthy Communities: Modeling Recreation Programs with STELLA

If you’re a 13-year-old who loves video games, Instagram and Cheetos, chances are someone – your parents, your teacher, the school nurse, Michelle Obama – is on your back to get up off the couch and be more physically active and socially engaged. They’re also encouraging you to eat nutritious food. They’re worried about your health; the scary probability that you’ll become obese and suffer from diabetes or heart disease or other chronic illnesses. But what if you don’t have the opportunity or support you need to orchestrate major lifestyle changes?

“Kids between the ages of 5 and 10 are active and engaged,” says David Compton, founder of the Healthy Communities Research Group and Professor Emeritus at the University of Indiana and University of Utah. “They’re encouraged by their parents and enrolled in programs that support physical and social engagement and offer snacks. But by age 10, increased athletic competition, the cost of uniforms and travelling teams, complex logistics and parental structures make it difficult or impossible for kids to participate. By age 15, skill requirements are even higher, parents are more disconnected or absent, and only the very best athletes are retained.”

That’s a problem for kids who really do risk their health when physical activity, social engagement and nutrition are traded for solitary, sedentary, digital endeavors. It’s also a huge problem for society that is paying a higher and higher cost for health problems associated with obesity.

“In 2007, about one-third of 10-to-17-year-olds were overweight,” says Kiboum Kim, Senior Research Associate at the Healthy Communities Research Group citing a 2010 study conducted by Barclay. “Fifty percent of obese children and 80% of obese adolescents remain obese as adults. If kids have two obese parents, they have an 80% chance of being obese themselves. That’s important to understand because as more obese adults become parents, it creates a reinforcing loop and there will be more obese children.”

...as more obese adults become parents, it creates a reinforcing loop and there will be more obese children

Kim also points out the costs – personal and societal – of obesity. “The annual medical costs encountered by an obese child are three times that of a healthy weight child. Obese adults spend $2,741 more on health care than their healthy weight peers. Obesity already accounts for 22% of medical spending. As obesity rates increase, so do associated medical costs. In fact, each year between 2001 and 2006, medical costs increased about 10% for the obese population and 6% for the overweight population. Costs will increase at even higher rates in the future.”

“The data shows real evidence of an obesity pandemic,” says Compton. “But, there are few coordinated efforts to get and keep young people socially and physically active. As we’ve looked at the data, we’ve come to understand the link between the increase in obesity and decrease in youth activity to be a system problem.”

It’s not surprising that Compton would immediately see the systemic nature of childhood obesity. “I started doing system analytics in the 1970s when it was crude,” he says. He introduced Kim, his student at the University of Utah, to system dynamics and STELLA. When Compton founded Healthy Communities Research Group, a multidisciplinary look at the relationship between parks and recreation agencies and health, Kim took up the task of creating a model of the system.

“We brought together anthropologists, educators, public health officials, exercise physiologists, nutritionists, community planners, and landscape architects to formulate a strategy for using parks to influence youth obesity rates,” says Compton. Now, Healthy Communities works in partnership with Green Play Research, Education, and Development (GPRED), an organization that helps public agencies create innovative programs that promote physical and social engagement in ways that improve individual and public health.

The Healthy Communities model creates a platform for cities, counties, and districts that seek to establish coordinated efforts to impact behavioral outcomes

Built using STELLA from isee systems, the Healthy Communities model creates a platform for cities, counties, and districts that seek to establish coordinated efforts to impact behavioral outcomes through systems analytics, database management, and ongoing surveillance of public health. The platform includes documented GIS documentation of built and natural assets – pools, playing fields, gyms, trails, open space, etc. – and data describing their service level quality, utility and proximity to youth populations.

“To that data we add an innovative layer of ‘affordances,’” explains Compton. “Affordance data catalogs services, programs, lessons, and activities that are available to the public by age, location, skill level, frequency, season, activity type, and other factors. They might be under the auspices of public parks and recreation departments or delivered in conjunction with contractors, associations, other non-profits or specialized service providers.”

Information about local policies, ordinances, laws and rules (like skateboard park or bike access restrictions) is added to the platform. When available and applicable, Healthy Communities adds data describing evidence-based practices that engage and sustain youth in active environments.

“The model is very complex and can be huge,” says Kim. “STELLA has made it easy to break the model into five modules that are consistent across public entities but unique in how they work together. Different agencies or organizations can tailor the model to reflect their use of funds, staff deployment, specific policies, or other variables.” Modeling how money is used for youth programs, funding sources and capital investment, population of 10 – 14 year olds, engagement in physical activity, and health care costs, the modules work together to form a surveillance and management system for a healthy community.

“We need to find ways to change habits of children to turn the obesity pandemic around,” says Compton. “We’re seeing that the deification of sport is a key factor in the physical activity and engagement drop-out rate. Parks and recreation agencies and the communities that fund them should consider redirecting funds from competitive, or consequential activities to non-competitive, or non-consequential activities for kids that want to participate for fun. And they need to understand and address other factors in the drop-out rate.”

Kim explains that the model shows how a diversion of funds from consequential to non-consequential activity impacts obesity rates and health care costs. It also illustrates how new approaches to youth recruitment and retention are needed. “Some agencies will have to change their culture,” says Compton. “They’ll have to learn how to treat eligible young people like customers.”

Storytelling helps illustrate the model to people who aren’t familiar with Systems Thinking

“It’s essential to use STELLA as a training tool,” says Compton. “Storytelling helps illustrate the model to people who aren’t familiar with Systems Thinking. It shows them that they need to do more than simply move people through the turnstiles. They need to document quality and retention to get results.”

The model is currently helping to identify youth program innovations in South Bend, Indiana, and Liberty, Missouri. Working through GPRED, Healthy Communities is reaching out to more public entities that fund facilities and certified youth activity programs. With 7500+ public parks and recreation agencies located across the US, the model offers administrators a systems analytics approach to managing obesity over the long term through surveillance and improved, evidence-based practices.

To learn more about Healthy Communities, visit Healthy Communities Research Group
David Compton — compton@indiana.edu, (801) 673-2907
Kiboum Kim — kimki@indiana.edu, 812-929-0363

An Open Standard for a Global Systems Thinking Market: XMILE Facilitates Model Sharing, Interoperability and Reuse

Anyone who has experienced the power of Systems Thinking to explain the behavior of a complex system, test solutions to big problems, and unite opposed stakeholders through data-driven exploration is left to wonder — why isn’t Systems Thinking more widely used? That’s the question Steve Adler, Information Strategist, IBM had after learning about isee systems tools and applications. He also knew the answer.

“When I first learned about System Dynamics and isee systems’ Systems Thinking approach I was surprised there were no technical standards that would enable people to, for example, share or reuse models made with different tools,” says Adler. “System Dynamics does have a standard nomenclature but it’s derived from an academic, ‘chalk board’ approach.”

isee systems did have a vision for a standard that would enable model sharing and reuse in the System Dynamics market. Karim Chichakly, Strategy Consultant, isee systems, wrote a draft of the standard in 2006 and isee systems has implemented that standard in both STELLA and iThink. “Steve said to us, ‘What about publishing an open source standard through a standards committee’ and when we understood the benefits outside of System Dynamics we were excited to go ahead,” says Chichakly.

and when we understood the benefits outside of System Dynamics we were excited to go ahead

In June 2013, the XML Interchange Language (XMILE) for System Dynamics Technical Committee convened and began to work on a standard that will open new markets by increasing model interoperability, portability, reusability, and ease-of-use. Co-chaired by Chichakly and Adler, the committee also includes members from Ventana, Forio, Mitre, Department of Defense, and the System Dynamics Society. Their refined draft standard will be reviewed and approved by the Organization for the Advancement of Structured Information Standards (OASIS), an international consortium that promotes the adoption of information formats including Open Data, OpenDocument, and DocBook.

Chichakly points out that current Systems Thinkers will benefit greatly from the new standard. “Systems Thinkers are always eager to share their models and a standard will allow them to collaborate across tools,” he says. “Models can be stored in cloud-based libraries, shared easily between organizations, and re-used to consider a wider range of options. ISVs will be able to develop new tools and apps quickly and, because XMILE is human as well as machine readable, Systems Thinkers will be able to do more on their own like generate custom reports and connect to graphics tools.” The standard also fits with the work flow of commonly used version control software to support collaborative model building.

With work on the standard underway, the group is eager to educate and excite a wider market about System Dynamics. “We need to show people the link between big problems, big data, and System Dynamics,” says Adler. “Ecosystems that exist in nature, business, health care, energy, social organizations, and many other areas are complex and hard to understand. It’s especially difficult to determine the cause of problems and the effect of potential solutions. By analyzing big data, people can find answers to big problems. Interoperability between data analysis platforms and System Dynamics tools will allow them to model and simulate the impact of those answers.”

We need to show people the link between big problems, big data, and System Dynamics

A webinar series, Big Data, System Dynamics, and XMILE, will showcase exemplar applications in the areas of environment, business, health care, and public policy. Jointly sponsored by isee systems, IBM and the OASIS XMILE Technical Committee, the series kicks-off on October 8 with The Dynamics of Climate Change: Understanding and Influencing the Planet's Future. “Everyone is invested in climate change and there are a lot of opinions,” says Adler. “Modeling provides a non-aligned way to look at a lot of data that answers those questions and then test policy scenarios.”

Andrew Jones, Co-Director of Climate Interactive, will introduce participants to C-ROADS, an award-winning computer simulation that helps people understand the long-term climate impacts of policies designed to reduce greenhouse gas emissions. World leaders are using the model in global climate negotiations and Jones will describe how it can be used by others to understand and test their own scenarios or conduct real-time policy analysis.

For more information about XMILE, read Karim Chichakly’s blog post XMILE – An open standard for system dynamics models.