Category Archives: Main Blog

New CODAP Website

Check out our newly revamped CODAP website!

Our new CODAP website design. https://codap.concord.org

Our newly designed and upgraded CODAP website has a new and fresher look. Visit us at codap.concord.org

We hope you like our new CODAP website, which is designed to make it easier to find information about CODAP and data science education. In particular, we hope you like the design which, is more modern and is intended to be easy to navigate.

We wanted a new website to better collaborate with our CODAP community, including educators, software developers, partners, researchers, and curriculum developers. You are now able to post comments, CODAP questions, and share use cases on our CODAP forums. We check the forum every day to keep in touch with you about all the things happening related to what we’re most passionate about—helping our users work with each other in the emerging field of data science education!

In addition to our new website, we invite you to join us for our upcoming webinar series starting on May 30.

Our first speaker is Cliff Konold of SRRI Education. Cliff will lead a discussion entitled “Modeling as a core component of structuring data” based on an upcoming article (Konold, Finzer, & Kreetong, in press), which describes research on student understanding of and ability to organize complex data for analysis. We expect lively discussion with Cliff’s facilitation.

Please RSVP for our May 30 webinar using our Eventbrite page here.

We are continuing to update our help pages and example documents with useful information about our work and also hope to include useful information on data science education related issues, so please do check back with us for updates.

Please contact us at codap@concord.org to let us know what you think of our new website—all comments and feedback are welcome. Please also let us know if you cannot find something or would like to make any suggestions for new information or topics.

Many thanks for your ongoing support and we look forward to hearing from you!

Designing ground-mounted solar panel arrays: Part III

Fig. 1: Rows of solar panels on racks in a solar farm
The most common configuration of solar farms is perhaps arrays consisting of rows of solar panel racks such as shown in Figure 1. But have you ever thought about why? Can we challenge this conventional wisdom?

Fig.2: Cover the field with horizontally-placed solar panels
Obviously, some inter-row spacing allows for easier cleaning and maintenance and, perhaps, even integration with agricultural farming (e.g., growing mushrooms that prefer shaded areas). But let's put those benefits aside for now and just consider the energy part of the problem. Let me point out a fact: If we completely cover the entire field with solar panels with zero tilt angle and zero gap (Figure 2), we are guaranteed to capture almost every single photon that strikes the area regardless of time and location. Such a simple-minded "design" will produce the maximal output of any given field at any location and time and there is absolutely no such problem as inter-row shading. So what solar design?
Fig. 3: Comparing two hypothetical fields.

It turns out that, although the simple-minded design can surely generate maximum electricity, each individual solar panel in it does not necessarily generate a maximum amount of electricity over the course of a year, compared with other designs. In other words, it may just use more solar panels to generate more electricity. As engineering design must consider cost effectiveness and even put it as a top priority, an engineer's job is then to look for a better solution that maximizes the production of each solar panel.

Fig. 4: Compare outputs of single panels in two fields (Boston).
A great advantage of Energy3D is that it allows one to experiment with ideas rapidly. So let's create a field with tilted rows of solar panels and leave some gap between them and then use the Group Analysis Tools to compare the daily and annual outputs of individual solar panels in the two hypothetical fields (Figure 3). And let's assume the fields are in Boston.

Fig. 5: Compare outputs of single panels in two fields (Phoenix).
Figure 4 shows that the total annual output of a single solar panel in the field of tilted rows is nearly 20% higher than that of a single solar panel in the field of flat cover in Boston (42° N). In this simulation, the tilt angle was set to be equal to the latitude. This cost effectiveness is one of the main reasons why we choose tilted rows of solar panels in high-latitude areas (aside from the fact that tilted angles allow rain to wash panels more efficiently and snow to slide from them more quickly).

Caveat for low-latitude locations


Fig. 6: Compare outputs of single panels in two fields (Mexico).
Note that this result applies only to high-latitude areas such as Boston. If we are designing solar farms for tropical areas such as Singapore, the story may be completely different. In low-latitude areas, small or even zero tilt angles make sense. Therefore, the design principle may be to cover the field with as many solar panels as possible or to use trackers to increase individual outputs (whichever is more economic depends on the relative prices of solar panels and solar trackers that change all the time). You can experiment with Energy3D to find out at which latitude this principle starts to become dominant. Figure 5 shows that the results in cities with a lower latitude such as Phoenix (33° N) and Mexico City (19° N) in North America. In the case of Phoenix, AZ, the gain from the tilted rows drops to about 10%. In the case of Mexico City, it drops to 5%. So designing a ground-mounted solar array for Mexico may be very different from designing a ground-mounted solar array for Canada.

National Science Foundation funds research and development of an IoT platform for smart schools

Fig. 1: A schematic illustration of IoT as a STEM learning integrator
Future sustainable and resilient infrastructure is expected to be powered by renewable energy, be able to respond intelligently to changes in the environment, and support smart and connected communities. We are pleased to announce that the National Science Foundation (NSF) has awarded our team a $2.9 million, four-year grant to explore the STEM education and workforce development challenges and opportunities in the coming transformation of our nation's infrastructure.

One of the core innovations will be a cyber-physical engineering platform for designing Internet of Things (IoT) systems that manage the resources, space, and processes of a community based on real-time analysis of data collected by various sensors. This innovation is potentially transformative as it can turn the entire building of a home, the entire campus of a school, or the entire area of a town into an engineering laboratory with virtually unlimited opportunities for learning, research, and exploration.

Fig. 2: A possible IoT system for managing a parking lot
Designing an IoT system provides plenty of opportunities to learn math, science, engineering, and computation practices in an integrated fashion, rather than in isolation. Working with sensors allows students to learn the science behind them through inquiry. For example, to calibrate an IoT system, students must understand what specific variables the sensor data represent scientifically. They must analyze the data to explore in what ranges the variables are supposed to vary in different scenarios in order to determine which type of response should be triggered, to what, and when. The acquired knowledge is then applied to the design of an IoT system, which requires engineering design thinking to make trade-off decisions, optimize system performance, and achieve cost effectiveness. Finally, the control, response, and integration of the entire system are realized through computer programming that deals with all foreseeable complexities. The overlaps among three basic skills—scientific reasoning, design thinking, and computational thinking—supported by the IoT platform provide researchers an opportunity to study their integration, as illustrated in Figure 1. (In fact, mathematical thinking is also involved, but let's just leave that out for now.)

This project is unique to engineering and computer science education because IoT is not only a crucial part of electrical engineering and information technology, but it is also one of the few ways through which computer programming can be directly linked to scientific inquiry and engineering design in the material world. Figure 2 provides an example.

This work is supported by the NSF under grant number 1721054. Any opinions, findings, and conclusions or recommendations expressed in this paper, however, are those of the author(s) and do not necessarily reflect the views of the NSF.

Artificial intelligence research for engineering design

Have you ever thought about what a pity it is when a senior engineer with 40 years of problem-solving experience retires? Have you ever thought about what a loss it is when a senior teacher with 40 years of teaching experience retires? Imagine what we could do for humanity if we find a way to somehow preserve their experience, expertise, and intelligence automatically before these incredible treasures are taken to the graveyard...

Heat map visualizations of different patterns of design task transition
Funded by the National Science Foundation, I have been working on the research and development of artificial intelligence (AI) for engineering design for a number of years and have been developing the Visual Process Analytics for visualizing and analyzing engineering design process data. This exciting intersection among AI (basically everything about how intelligence can be realized), engineering (basically a generative and creative discipline), and cognitive science (basically everything about how humans acquire intelligence) is full of tremendous challenges, but it also creates unprecedented opportunities that constantly entice and enlighten me.

I have recently written a short article to explain my research to the lay people (mostly educators, but the implications are not limited only to education). Check it out at http://energy.concord.org/~xie/papers/aired.pdf

Data Science Education Meetup at Cyberlearning 2017

Thank you to the fantastic crowd at Cyberlearning 2017 who attended our Data Science Education Meetup at Mussel Bar and Grill on Tuesday night. The Meetup provided an opportunity for members of the Cyberlearning community to discuss innovative ways that we could implement data science education into research and curriculum development, to address NSF’s 10 Big Ideas for Future Investments strand in data science [PDF], and to enjoy some great food!

Our attendees at the Data Science Education Meetup at Cyberlearning 2017

Attendees at the Data Science Education Meetup at Cyberlearning 2017

There were several strands of activity toward data science education at Cyberlearning 2017, and the Meetup provided an opportunity for us to share and reflect on the following:

  • A roundtable session on Teaching Data Science, in which our own William Finzer facilitated a discussion on our Data Science Education Technology conference in Berkeley in February.
  • Hearing about some of the challenges researchers face in implementing curriculum at the middle school and high school level, as well as some of the Cyberlearning projects (such as Data Science Games, CODAP, Impact Studio, Learning and Youth, and STEM Literacy Through Infographics), which are hoping to address these challenges.
  • Participating in the Data Science Education for 21st Century Learning CL17 Working Sessions Strand. It was especially great to work with Sayamindu Dasgupta and Jesse Bemley to identify the important aspects necessary to move data science education forward.
  • Identifying data science as part of computational thinking, as put forward by NSF Program Officer Arlene M. de Strulle on Wednesday morning.

Our Meetup was packed (we had to move to a larger set of tables, as a matter of fact!), and the evening was full of lively discussion about the impact and need for data science education. We thank everyone who attended and are especially grateful to our community for bringing in so much passion and energy for data science education.

Please check out our future Data Science Education Meetup dates and locations at concord.org/meetup. Or join us online through a series of monthly Data Science Education Webinars, starting in May. We have a great lineup of speakers, including Cliff Konold, Amelia McNamara, and Rob Gould. We’ll post dates and additional information at concord.org/meetup.

As always, please leave a comment or suggestion for future Meetups, community-building activities, or future innovations in data science education you’d like to see. We love hearing from you.

A demo of Infrared Street View

An infrared street view
The award-winning Infrared Street View program is an ambitious project that aims to create something similar to Google's Street View, but in infrared light. The ultimate goal is to develop the world's first thermographic information system (TIS) that allows the positioning of thermal elements and the tracking of thermal processes on a massive scale. The applications include building energy efficiency, real estate inspection, and public security monitoring, to name a few.
An infrared image sphere


The Infrared Street View project is based on infrared cameras that work with now ubiquitous smartphones. It takes advantages of the orientation and location sensors of smartphones to store information necessary to knit an array of infrared thermal images taken at different angles and positions into a 3D image that, when rendered on a dome, creates an illusion of immersive 3D effects for the viewer.

The project was launched in 2016 and later joined by three brilliant computer science undergraduate students, Seth Kahn, Feiyu Lu, and Gabriel Terrell, from Tufts University, who developed a primitive system consisting of 1) an iOS frontend app to collect infrared image spheres, 2) a backend cloud app to process the images, and 3) a Web interface for users to view the stitched infrared images anchored at selected locations on a Google Maps application.

The following YouTube video demonstrates an early concept played out on an iPhone:



Data Science Education Meetup at NSTA 2017

Thanks to all the great folks who attended our NSTA 2017 Data Science Education Meetup at BottleRock LA last night. We had a great crowd attend, complete with representatives from Magnitude.io, the education arm of the International Space Station (ISS), CASIS, the folks from MiniPCR, LAUSD, Lodi USD, the CREATE for STEM Institute, Educational Passages and more! We’re still processing the yummy food and hours of great conversation about all things data.

Happy attendees at the Data Science Education Meetup at NSTA 2017

A few highlights of the evening included:

  • Hearing about streaming atmospheric data from Magnitude.io’s successful rocket and balloon launches at Lodi schools and others all across California, and their ExoLab experiments on the ISS
  • Debating the grand challenges and barriers to data science education with teachers and district staff
  • Hearing about how using data, modeling and evidence has transformed teaching for those using the Interactions curriculum from the CREATE for STEM Institute and the Concord Consortium
  • Seeing Liam Kennedy’s amazing ISS-ABOVE device that changes your TV into a portal to the International Space Station, complete with live alerts when the ISS passes overhead
  • Learning about how MiniPCR’s Genes in Space competition is bringing high school students’ experiments to the space station and how a new USB-stick-sized gene sequencing device is speeding up microbe research on the ISS by orders of magnitude
  • Hearing stories of how data from GPS-tracked drifter boats from Educational Passages has connected students across continents
The place was packed and the evening was full of great geekery, all the more evidence that the time has come for data science education. We thank everyone who attended and we look forward to building networks, collaborations and new modes of teaching and learning together.
Missed us last night? No problem. Join us for the next one, at an upcoming conference near you! See all our meetups and RSVP at concord.org/meetup.

Global Researchers and Developers Connect at 2017 Data Science Education Technology Conference

Over 100 thought leaders from organizations around the U.S. and four continents gathered from February 15 to 17 to generate important innovations needed in technology and teaching and learning at the Concord Consortium’s first Data Science Education Technology conference. Senior researchers, educators, and scientists from as far as Nigeria and New Zealand convened at the David Brower Center in Berkeley, California, to share a comprehensive overview of this rapidly growing area of data science education.

“This is a thrilling milestone,” notes Chad Dorsey, president and CEO of the Concord Consortium. “Never before have experts across mathematics, science, and technology come together to focus their thinking on the emerging field of data science education.”  

Dan Damelin, William Finzer, and Natalya St. Clair of the Concord Consortium.

Some of the 100 DSET Conference attendees.

Over the next few decades, data science education is expected to undergo a profound change. According to opening panelist Deb Nolan, UC Berkeley is offering an Intro to Data Science Class with enrollment of over 700 students in it, and there is currently a campus-wide initiative to plan for a data science major. In addition, the National Science Foundation has included Harnessing Data for 21st Century Science and Engineering as a priority in a “10 Big Ideas for Future Investments” report.

To reflect the dual-sided nature of this new territory in education, the conference topics comprised two strands. The Teaching and Learning strand, designed for those considering data science-related curriculum development, brought together researchers and educators, and generated a wealth of new understandings and patterns for educational research—educators in particular left with fresh ideas on how to apply cutting-edge curriculum practices in their classroom. Researchers left with important ideas about how to define success for educational results in this fast-growing and essential research field.

“Data science is evolving. And data science education is a science on its own, which can reach out to all branches of science and similar social sciences. Already I see people here who are trying to brainstorm ways we can move ahead on data science and how we can move forward to educate people on data science,” says Kenechukwu Okeke of Nnamdi Azikiwe University, Nigeria.

In the Technology strand, participants worked with data analysis platforms such as Tuva, Desmos, and CODAP (the Common Online Data Analysis Platform). Participants with software experience created their own CODAP plug-ins and also built skills to make their work more efficient. In the closing session, DSET participants watched as Ph.D. student Takahiko Tsuchiya (Georgia Institute of Technology) made data in CODAP audible as sound using a sonification plug-in he had programmed during the conference.

Register now for CADRE Webinar on DSET conference highlights

Join William Finzer, Daniel Damelin, and Natalya St. Clair of the Concord Consortium in a CADRE webinar Friday, March 3, from 1-2 pm EST for highlights from the DSET conference. Register now!

Join us at future DSET Meetups

We’ll continue to gather thought leaders in data science education technology through informal meetups in Austin, San Antonio, Los Angeles, Portland, Baltimore, and New Zealand. Going to SXSW Edu, NCTM, or NSTA? Contact us for more information about upcoming meetups. We look forward to seeing you!

 

Designing building-integrated photovoltaics with Energy3D

Fig. 1: An example of solar facade.
Building-integrated photovoltaics (BIPV) represents an innovative way to think and design buildings as both human dwellings and power plants. In BIPV, solar panels or photovoltaic thin films are used to replace conventional constructional materials in parts of the building envelope such as roofs, walls, and even windows. Designing new buildings nowadays increasingly includes BIPV elements to offset operational costs. Existing buildings can also be retrofitted with BIPV (e.g., replacing glass curtain walls with solar panels). BIPV is expected to grow more important in architectural design and building engineering.

Fig. 2: An example of solar curtain walls
We are developing modeling capabilities in Energy3D to support the design, simulation, and analysis of BIPV. Figures 1 and 2 in this article show a few cases that demonstrate these capabilities in their primitive forms. Considering BIPV is relatively new and a lot of research is still under way to develop and test new ideas and technologies, we expect the development of these capabilities in Energy3D will be a long-term effort that will be integrated with latest research and development in the industry.
Fig.3: Power balancing throughout the day.

As the first step towards that long-term vision, the current version of Energy3D has already allowed you to add solar panel racks to any planar surface, being it horizontal, vertical, or slanted. Running a simulation for any day, you will be able to predict the daily output of all the solar panels. You can also compare the outputs of selected arrays. For example, if you want to track down on which side solar panels produce the most at a given time during the day, you can compare them in a graph. Figure 3 shows a comparison of the solar arrays in the model shown in Figure 1. As you can see, the east-facing array produces peak energy in the morning whereas the west-facing array produces peak energy in the afternoon. In this case, the BIPV solution ensures that the photovoltaic system generates some electricity at different times of the day.

Video tutorial: Solarize an imported building in Energy3D

We are pleased to announce that the solar panel and analysis tools in Energy3D (version 6.5.6 or higher) are now fully applicable to arbitrary imported structures. We hope that the new capabilities can help engineers who design rooftop solar systems and building solar facades to get their jobs done more efficiently and students who are interested in engineering to learn the theory and practice in an inquiry-based fashion. The six-minute video in this article demonstrates how easy it is to perform solar panel design and analysis in Energy3D. (Note: Unfortunately, the annotations in the video do not show if you are watching this on a smartphone.)



One of the handiest features is the automatic, real-time detection of the angle of the surface under a solar panel while the user is moving it. This feature basically allows the user to drag and drop a solar panel or a solar panel rack anywhere (on top of roofs, walls, or other surfaces) without having to set its tilt angle manually.

Solar heat map of a house with solar panels
Copy and paste a house with solar panels
Solar panels are "first-class citizens" in Energy3D as they are readily recognized by the built-in simulation engines. Energy3D provides a comprehensive list of properties that you can choose for each solar panel or solar panel array. For example, even the temperature coefficient of Pmax, a parameter that specifies the change of solar cell efficiency with regard to ambient temperature change, is supported. The software also has a variety of analytic tools for predicting the hourly, daily, and annual outputs of each solar panel and their sums. Interactive graphs are available to intuitively show the trends and allow the user to compare the outputs of different solar panels, of different arrays, on different days, or with different environmental settings (e.g., with or without a tree nearby).

These "native" solar panels are now completely blended into the "alien" meshes of structures imported into Energy3D from other CAD software or Google Earth. For example, once you drop a solar panel on a surface of the structure, it will stick to it. In other words, if you move or rotate the structure, the solar panel will go with it as if it were part of the original design. When you copy and paste the entire building, the solar panels will be copied and pasted as well (by the way, it takes only four clicks to copy and paste a building in Energy3D through the pop-up menu: one click to pick which one to act upon, one click to select the "Copy" action, one click to pick where to paste, and one last click to select the "Paste" action). That is to say, the native and alien meshes are completely meshed.