The Steinbeck Innovation Project: Applying Systems Thinking and Action Research to Regional Planning

While most corners of the U.S. are rebounding from the Great Recession of 2008, Salinas, California’s economy has remained moribund. The state’s water shortage is also a concern for the County-wide $4.14 billion agricultural industry that serves as the city’s economic foundation. A significant portion of the public school student population is first or second generation Americans; many of them struggle with English writing skills and cultural differences that discourage their academic success. Community security is threatened by a gang problem. It’s a community in need of an opportunity.

Salinas has everything you need to create a new model for sustainable economic and educational development

The search for an opportunity led the Mayor of Salinas to Captain Wayne Porter, Chair of Systemic Strategy and Complexity, Global Public Policy Academic Group at the Naval Postgraduate School (NPS) in Monterrey, California. A PhD candidate and Chair for Systemic Strategy and Complexity, Porter had previously served as the Special Assistant for Strategy to Chairman of the Joint Chiefs of Staff, Admiral Mike Mullen and had been sent to NPS to pursue the application of system science to strategic planning.

Porter recognized the challenges facing Salinas as an opportunity in disguise. “Salinas has everything you need to create a new model for sustainable economic and educational development,” says Porter. “It has an established industry and a young, available workforce. It has proximity to Silicon Valley where there is technological expertise and capital.”

The city was already at carrying capacity for water usage... It couldn’t build the economy by adding more cultivating acreage

His idea was to develop a regional cluster that would focus on developing and manufacturing the elements of precision agriculture – sensors, robotics, nutrient recovery and filtration systems, water management and waste management technology. “The city was already at carrying capacity for water usage,” says Porter. “It couldn’t build the economy by adding more cultivating acreage. But it could attract businesses and jobs that would leverage its agricultural base by developing a skilled, design and manufacturing labor pool with an exportable commodity.” The vision was to address the global challenges of food and water sustainability through the tools of precision agriculture while providing local opportunity and hope.

Together with the Mayor, Silicon Valley consultant John Hartnett and a small team of community stakeholders, Porter participated in the Steinbeck Innovation Project, a strategic planning and investment effort that launched the Steinbeck Cluster in December 2012. At the same time, he began the Systems-Thinking-fueled action research that would serve as his dissertation focus.

The systems model runs all five modules over a 25-year period and is initialized with publicly available data

“The Salinas project was perfect for action research and I was fortunate to be embedded in the process,” says Porter. “It allowed me to model socioeconomic, cultural, and physical resource restraints that you’ll find at a regional, federal, or even worldwide level. And, it was a great opportunity to apply Systems Thinking and strategic planning while helping the community of Salinas and the Central coast region.”

Using iThink from isee systems, Porter developed five, integrated modules that represent key systems with the project’s strategy. While not intended to be predictive in nature, the system model provides decision and policy makers an enhanced understanding of potential non-linear behavioral outcomes that could result from the strategy’s integrated structure and feedback mechanisms over time. The systems model runs all five modules over a 25-year period and is initialized with publicly available data (population, revenue, flow from high school to agriculture technology education, gang membership, water supply, etc.). The five modules that comprise the model address:

• Investment funds, start-ups, and research activity
• Gang membership and prevention programs
• Education for agricultural technology employment
• Water management
• The Salinas area’s attractiveness to high income earners

“The attractiveness module pulled everything together,” says Porter. “It looks at the ability of Salinas to attract agriculture technology professionals from five neighboring cities as a function of six factors of community attractiveness: community security, school quality, available services and infrastructure, cost of living, employment opportunities, and water availability. Each factor is given a weighted score. If the ratio of Salinas to the average score of the five cities is zero on any factor in a given year, it’s not attractive.”

The model has provided several critical insights... At the 19-year mark, according to the model run, projected growth will exceed water availability

Scored attractiveness factors allow Salinas to compare itself to nearby communities and anticipate movement of higher-wage earners. The model also points out the relationships between factors. “As higher-wage earners move in and business grows, tax revenue increases,” says Porter. “That leads to an improvement in school and service quality and community security. However, the cost of living also increases, diminishing attractiveness. Water is another balancing loop: as the community grows, commercial and residential consumption draw down the water supply.”

“The model has provided several critical insights,” says Porter. “Most of Salinas’ water is drawn from aquifers. At the 19-year mark, according to the model run, projected growth will exceed water availability. So, the city knows now that it will need a water sustainability solution well before that point. And, the thing that increases the number of sustainable businesses in Salinas is increasing the percentage of kids that get involved in Ag Tech, science, technology, engineering, and math (STEM).”

The gang membership module shows a decline in gang membership, which then influences several factors in the attractiveness model

The insight about kids’ involvement in STEM has been particularly powerful for planners and investors. “Tailored Ag Tech involvement makes more of a long term difference than increasing financial investment,” says Porter. “Further, STEM involvement inoculates kids from gang membership.”

Accordingly, the Steinbeck Innovation Foundation is supporting a program called Coder Dojo that gets kids involved in programming when they’re 8-13. By creating job opportunity and reducing the susceptible pool of 14-24 year olds through early STEM intervention, the gang membership module shows a decline in gang membership, which then influences several factors in the attractiveness model.

There’s no specific recipe for sustainability

“I’ve spent two and a half years looking at industrial clusters and there’s no specific recipe for sustainability,” says Porter. “There is, however, one consistent component that’s essential for success: an underlying educational system that exposes kids to the information and skills needed to create an adaptable workforce. The Steinbeck Cluster had to address economic development by building opportunity through education.”

Porter’s work extended beyond modeling the industrial cluster’s impact on the Salinas community. As important, he wanted to understand the impact of supporting the cluster’s development with strategic planning and System Dynamics. To do that, he interviewed project participants to understand the group’s strategic planning process. Then, he administered a short Likert questionnaire in which respondents rated aspects of their strategy and planning process. Subjects then interacted with the integrated systems model. They were encouraged to change parameters and note the impacts of those changes. Finally, the questionnaire was re-administered to note changes in thinking about the strategy and planning process.

If you’re trying to create a new policy, a new process, a new educational model; this is a decision maker’s tool and a great ‘what if’ tool

“It’s so validating to watch the people involved interact with the model,” says Porter. “They really get it for the right reasons. They focus on the feedback mechanisms and understand that it’s the change in relationship between numbers – not the numbers themselves - that shows key points of influence.”

When Porter shared the model with Bran Fitzgerald, an early Apple employee, high tech entrepreneur, and COO for a small Silicon Valley technology firm, it “blew his mind.” “If you’re trying to create a new policy, a new process, a new educational model; this is a decision maker’s tool and a great ‘what if’ tool,” says Fitzgerald. “It’s an innovator’s dream.”

To have a tool that is not just easy to use but that brings clarity to how to get to a solution efficiently... you can’t calculate the worth of something like that

A newcomer to Systems Thinking and modeling, Fitzgerald appreciated iThink’s accessibility. “It’s very easy to follow, very interactive,” he said. “When you start working with it, the results are very clear, which helps you set a course for where you want to go and how you want to get there. In the technology industry, time isn’t just money. Time is important. To have a tool that is not just easy to use but that brings clarity to how to get to a solution efficiently... you can’t calculate the worth of something like that.”

“I’m really hoping more and more people catch on to the benefits of Systems Thinking and modeling,” says Porter. As he hoped, his model is getting attention from peers at NPS and the Pentagon. Maybe more important, it’s turning the heads of young Salinas students who will use it to plot their own futures.

“When students see this they become zealots,” he added.

How much water is down there? Tracking groundwater systems with STELLA

Local authorities around the world share a worry – water. As populations grow and the climate changes, answering the questions “How much water do we have now?”, “How much water will we have over time?” and “Will it be enough?” is central to ensuring community viability. As straightforward as they sound, those are difficult questions to answer.

Could a systems analysis approach using models and simulation be used to understand a defined ground water system?

Professors at the Newcastle University School of Civil Engineering and Geosciences were particularly interested in finding ways to answer those questions for communities reliant on ground water resources. Systems analysis is often used to understand surface water systems but aquifers are often a black box in those analyses. The domain of ground water models that does exist requires a lot of data and takes a specific approach that doesn’t fit all aquifers. Could a systems analysis approach using models and simulation be used to understand a defined ground water system?

It was a perfect Master of Science dissertation project for Andreas Markou. “My career has been in software development and I’ve always been passionate about geoscience,” says Markou. “This project brought those two halves of me together.”

Markou focused his work on The Fell Sandstone aquifer, the sole water source for the Berwick upon Tweed area in far northern England. Understanding the ongoing condition of the aquifer is a central interest of the water company, national regulation agency and local planning authorities.

STELLA lets you be very productive in thinking about a problem without having to delve into coding... It frees you up to look at the big picture

“The Fell Sandstone aquifer is geologically complex and there’s not much data to describe it so available models and approaches wouldn’t work,” says Markou. “My dissertation was an exploration of a system dynamics approach to ground water modelling. The outcome could have been ‘We tried using system dynamics but it didn’t work.’ I moved softly into the project and created a series of models that represent an exploratory path.”

Markou created his models using STELLA from isee systems. “STELLA lets you be very productive in thinking about a problem without having to delve into coding,” says Markou. “It frees you up to look at the big picture. You can draw models or diagrams with other solutions but STELLA allows you to think graphically using an intrinsic rule set that includes stocks and flows. And you can sketch out problems at a map or causal diagram level.”

Building those models helped me gain confidence with modeling and graphical functions

Causal diagrams were the basis of Markou’s initial exploration, a high-level look at water supply and demand. A first simple model includes two head values at hydraulically linked points and simulates system behaviour over time when flow is the only influence in place.

That was followed by more evolved forms of the model that used several years of available data to demonstrate bore hole levels during pumping tests and abstraction. “Building those models helped me gain confidence with modeling and graphical functions,” says Markou

Next it was time to understand how the aquifer worked with inputs and outputs over time. Markou studied the overall geometric volume of a particular aquifer block and the surface area that would support its recharge. “It’s a bucket concept,” says Markou. “It doesn’t require a lot of GIS data but allows us to consider rainfall and catchment versus usage.” The model made it clear that direct recharge – immediate soil infiltration - would not balance the aquifer at the required levels. In fact, the aquifer would require as much indirect recharge – stream recapture - as direct to remain at equilibrium.

We could see what would happen if rainfall increased by 50% or if extraction increased by a certain amount

Next it was time to understand how the aquifer worked with inputs and outputs over time. Markou studied the overall geometric volume of a particular aquifer block and the surface area that would support its recharge. “It’s a bucket concept,” says Markou. “It doesn’t require a lot of GIS data but allows us to consider rainfall and catchment versus usage.” The model made it clear that direct recharge – immediate soil infiltration - would not balance the aquifer at the required levels. In fact, the aquifer would require as much indirect recharge – stream recapture - as direct to remain at equilibrium.

“I created a dashboard that allowed ‘What if’ testing,” says Markou. “We could see what would happen if rainfall increased by 50% or if extraction increased by a certain amount. That allowed simulation of the fall and rebound of water in the aquifer. We could see when levels would rebound and when spring flows would start.”

Spatial mapping, which is important in understanding geological systems that change in composition and porosity from point to point, does require in-depth GIS data. While Markou did not have access to the level and quantity of data required for intensive spatial analysis, STELLA Spatial Map did prove capable of modelling localized behaviour around a borehole.

Findings helped explain an optimistic view by stakeholders of the amount of water that could be taken out of the system during early operation

“The study and models produced a number of findings that are very supportive of already held conclusions,” says Markou. “For example, it was believed that the aquifer has better than average storage and flow because of fracturing and the model supports that view.”

The models also indicated that some parts of the aquifer are relatively discrete with geological barriers between them. Both findings helped explain an optimistic view by stakeholders of the amount of water that could be taken out of the system during early operation.

“The scientific community is very demanding in terms of the amount of data behind conclusions,” says Markou. “Even though there wasn’t much data for this aquifer, there were still real questions being asked by the stakeholders. I showed that whilst the models can’t offer total precision, they can provide insight into the behaviour of the aquifer.”

STELLA’s graphical function tools that handle time series data are impressive

“STELLA’s graphical function tools that handle time series data are impressive,” says Markou. “I could draw the shape of data even when I was missing some numbers. Sensitivity tools allowed me to funnel down on a specific variable even when I didn’t know what the exact data values were.”

The models produced during Markou’s exploration of a system dynamics approach to understanding ground water supply are in the hands of the incumbent water company. The regulatory agency has received his full report and been very supportive throughout the study. The visual nature of STELLA lends itself well to projects requiring multi-party collaboration.

STELLA provides the perfect glue for looking at integrated water and ecology systems

While his dissertation work was Markou’s first in-depth use of Systems Thinking and STELLA, it won’t be his last. “My number one priority is to be a system dynamics practitioner,” he says. “I’m already working on a project that looks at the same sort of water system but adds a microbiological twist – a bug that could impact water quality. STELLA provides the perfect glue for looking at integrated water and ecology systems.”

Markou is also hoping to use STELLA and Systems Thinking for organizational projects but he remains drawn to geoscience. “Through my dissertation work I discovered a number of geological topics that I want to get back to with STELLA,” he says. “I can develop software for my work but it’s a pleasure to have a tool that works at such a high level and keeps you focused on the problem rather than the code. The creative part of your brain really takes off.”

From Systems Thinking to Dynamic Modeling

Don’t have the confidence to use STELLA or iThink to tackle big issues? We want to help! In March, 2014, isee systems started a new series of online courses designed for anyone who is interested in learning how to use STELLA or iThink for building models - no prerequisites required! Our software was designed to simplify the modeling process and to clearly communicate the dynamics of systems to anybody. In the same vein, this new series, From Systems Thinking to Dynamic Modeling, is equally as accessible and truly valuable in helping our customers become proficient in modeling and explaining the impact of dynamic relationships.

Customers who attend the live classes will be able to ask questions, get feedback on homework, and participate on the class forum. However, because not everyone can easily attend live classes due to scheduling conflicts or different time zones, each course is recorded and will be available for sale on our web site. As the courses are online and recorded, you will be able to learn whenever and wherever you want.

Each of the six courses in the series includes four one-hour classes. The first four courses build on each other while the last two include assorted, intermediate topics. Once a course is purchased, all class materials including a recording of the webinar, homework assignments, PowerPoint slides, models, and any supplemental reading materials will be added to each participant’s account shortly after the class takes place. Like all of our online training these courses will never expire or disappear from your account, so you can continually return to them to brush up on your modeling skills.

The first two courses build the foundation, starting at the beginning with an overview of concepts that are helpful when setting out to build a model. The building blocks, rules for constructing, and model behavior progress to making simple models. The next two courses expand on the first two and include common model structures and techniques used in creating successful models. The last two courses will cover intermediate topics such as arrays, conveyors, and discrete simulation.

Whether you start at the beginning or jump in somewhere in the middle, we think you’ll find this series of courses helpful in getting the most out of your STELLA or iThink software!

On the Road

On June 28th to 30th, isee systems will be once again participating in and attending the Systems Thinking & Dynamic Modeling Conference for K-12 Education, hosted by the Creative Learning Exchange in Wellesley, Massachusetts!

We are ecstatic to be sharing our software and books with educators and continuing to help introduce Systems Thinking to future generations. Co-President Bob Eberlein will be giving a workshop with our newest isee product, STELLA Modeler® for iPad. If you are interested in learning more about our workshop or the conference please visit www.clexchange.org.

Then from July 20th to 24th, isee systems will be sponsoring and attending the 32nd International Systems Dynamics Conference in Delft, The Netherlands!

We are excited to exhibit our software and books while having the added benefit of meeting many of our fellow Systems Thinkers in person. We will be on hand to answer questions about your modeling projects and how isee systems can help, as well as listen closely to learn where we can improve both our products and services.

Get a sneak peak in our User Group Meeting, where we’ll be demonstrating our latest software developments, share product plans, and our directions for the near future. We will also host an Exhibitor Demo and facilitate a XMILE Roundtable discussion regarding the status of the new open standard and hear your comments.

Co-presidents Karim Chichakly and Bob Eberlein will conduct three workshops during the conference. They include Getting Started with STELLA, Organizing and Analysing Models with STELLA and iThink, and Building and Publishing Model Interfaces using STELLA and iThink. If you would like to learn more on these workshops visit System Dynamics Conference. We greatly look forward to connecting with the System Dynamics community, seeing familiar faces and making new friends. Hope to see you there!