The 2.1285e+10 pixels of 2011 (so far)

How do you visualize an entire year of photography in a single graphic? The following is a geeky art project I indulged myself in. It measures every pixel I have taken so far this year, showing the prevalent hues and volume of photos throughout the year. Each bar represents a week of images - there would be 52 if the year were over. The length of the bar represents the number of pixels of photography done in that week. The colors in the bar show the hues captured during that week.

The quantity trends are pretty obvious. The first bump in January is indoor shots from my signature annual party. I took relatively few photos in the spring when I was locked in the thesis cave. Photo quantity grew in the summer as I embarked on weekly kayaking and hiking trips. Peak pixel was September as I took a week to bike through France, capturing hundreds of images along the way.

Hue trends are also present. Winter times have a lot of brown and white. Green appears more as summer approaches. I might be fooling myself but I think I can even see autumn foliage. You might notice the appearance of a red kayak and a red backpack if you look carefully. I had hoped that my brilliant yellow kayak would show up in the summer photos, but the boat's colors get spread out among too many different hues to be noticeable.

How did I make this? MATLAB of course!

  1. Sort a 512-color RGB colormap with RGB2NTSC
  2. Load an image with IMREAD
  3. Reduce JPEG's truecolor space down to a tractable 512 colorspace with RGB2IND
  4. Bin all pixels into groups with IMHIST
  5. Lather, rinse, repeat. Sum all photo histograms, binned by date.
The trickiest part was figuring out how to sort the 3-dimensional RGB color space into a pleasing 1-d continuum. It turns out that I'm not the first to have this problem; the folks over at Visualmotive already investigated this and I agree with their preference for the YIQ colorspace.

In the histogram above, you can see blue peaks for the sky, green for the trees, and orange for the hiker's shirt. Black is prevalent since shadows are dark and some objects (the dog, the shirt) are also black. The most heuristically "wrong" thing about the 512-color reduction is that purple/magenta hues show up more often than you'd expect. There's a bit of lilac in the sky, and the same color often also shows up in wood and stone.

This was a fun project. I like using computation to come up with new ways to understand the world.


No, You Can't Have A (Fully-) Solar Car

Do the Math is a quantitative blog that looks at current issues in a back-of-the envelope fashion. The latest entry is a calculation of exactly what it would take to make a production solar car run. I once delivered much the same calculation to a bunch of undergrads in a policy course during my solar car building days. They were kind of bummed by the numbers.

As far as I can tell, the only real application for on-vehicle solar cells is powering a fan in the car to keep it cool during hot days. This reduces air conditioning load when returning to the vehicle after it has been parked for a while, ultimately saving fuel or battery charge. PV is just too expensive and low-power for anything else. If you want a real solar car, charge your EV from the roof array on your house instead of hauling around a bunch of fragile cells.

Thanks to Ned Gulley for the reference.

Image by the MIT Solar Electric Vehicle Team. I wish they had a good photo of my beloved Manta.


Smart Grid Assists Wind Integration: A Non-Scary Thesis Talk

Want to know how Hawai'i can run its grid more efficiently, harvest more wind, and be more reliable all while not actually using many demand resources? Want a low-jargon to learn about what I have been doing for my thesis, with a promise to use zero equations and only one incidence of the word "stochastic?" Now is your chance, my friends. I will be the featured speaker at SDM's Monday 14 November web seminar. Register here for free and don't forget to throw a few difficult questions my way.


Most Innovation Is Invisible

Neal Stephenson's recent talk on "Innovation Starvation" strikes a nerve in every engineer: we don't build anything anymore. With the end of high-visibility mega-projects like the space shuttle, it's an understandable notion. Another way to look at it is that the innovation of our era is incremental and invisible.

The telcos have invested billions to create a worldwide high-speed mobile data network. The only manifestation of this gigantic project is the occasional poorly-hidden tower disguised as a tree. In exchange, we are never lost, can always meet our friends at an event with no planning, are always informed, record or reference any memory, and can travel in unfamiliar places like a local.

Our electricity system is undergoing a seismic shift away from coal and toward natural gas. For the last decade, 90% of the new generation capacity in ISO New England has been highly-efficient, relatively low-carbon combined cycle gas turbines. If each of these replaced a coal plant, you're talking an avoided-carbon equivalent equivalent to a few hundred wind turbines. Cape Wind is a big, visible project with a high feel-good factor. But the invisible innovations in natural gas exploration that have made this cleaner fuel relatively cheap have had much more impact.

Even the military (which used to spontaneously generate battleships and bombers like aristotelian flies) is assembling its toys from loose networks of small parts. The drone that just executed Anwar al-Awlaki is a fragile model airplane connected to a bunch of satellites, a guy in a trailer in Nevada, and world-class intelligence gathering.

To a generation that grew up on glossy books showing us the Future in its flying-car glory, this is all unsatisfying stuff. Sure, practical supersonic transport would cut my flight duration to Europe by a few hours. But the ability to rent a bike in Boston and return it in Cambridge saves more time per year. There is plenty of innovation happening, Neal. You just need to look with different eyes.


Special Project: NREL

I've been waiting for a while to announce this one. Last fall I applied for an Innovative Research Analysis Award Program grant from the National Renewable Energy Lab. It took a while to get the paperwork squared away, but the award is now official. From the NREL website:

Power System Balancing with High Renewable Penetration: the Potential of Demand Response in Hawai'i

The State of Hawai'i has adopted an aggressive renewable portfolio standard of 40% renewable energy by 2030. Most system balance studies in Hawai'i have focused on grid assets such as spinning reserve or energy storage to provide electricity when generation from renewable resources changes unexpectedly. Demand Response (DR) is an alternate strategy in which the grid operator ensures system stability by managing select consumers' loads, such as changing air conditioner set points or turning off non-essential loads within the service area according to a pre-approved prioritization plan. Demand Response may provide a lower-cost solution to balancing intermittent supplies, enabling the State to achieve its goals for reduced energy dependence. This research will use time series data for demand, wind speed, and wind speed forecasts to identify the potential grid-value of Demand Response, as well as DR program design to meet the needs of both the electric utility and electricity customers. A unit commitment model will simulate the relative production of wind, thermal, and demand response resources, then predict the frequency, duration, and scope of curtailment events necessary to maintain a balanced grid. Lessons learned in Hawai'i can be applied in other regions.

Collaborators: Massachusetts Institute of Technology, National Renewable Energy Laboratory

Estimated completion is September, with publications and conferences to follow. This project is the reason why I delayed my graduation from Summer to Fall. It's exciting to see it come together.


Meta: Plus

I now have a g+ identity. Use my profile to add me to your circles and I'll return the favor. Postings there (and still on facebook) will mostly be mirrors of what you see here.


It's Your (Grid) Weather Forecast

I'm a bit of a grid geek. We are having a hot day here in New England and my facebook/google+ stream is drowning in people talking about 100+ degree temperatures. That's the obvious result of the weather. The less obvious result is that electricity demand is currently surging as everyone cranks the air conditioning to stay comfortable. Almost every generation asset in the region is probably running near maximum right now. Since taking Ignacio's grid regulation class, I have recreationally checked the marginal price map from ISO New England. Today is the worst I have ever seen it. LMPs are in the $200/MWh range right now. During last week's mild summer temperatures the region was about $30/MWh.

What does this mean behind the scenes? Suppliers of base load power are cleaning up right now. If you run a cheap coal plant, you get paid the same as the near-decommissioned fuel oil plant that they crash fired yesterday. If they didn't have long-term power supply hedges, utilities would be losing money like crazy. As a residential customer, NSTAR charges me 7.7 cents/kWh to buy electricity and they could be paying 20 cents right now. Our independent system operator is no doubt going crazy to make sure that all the reliability constraints are being met. EnerNOC is probably calling industrial consumers all over the region and asking them to curtail their electrical load. This is all heroic effort to make sure that we all stay comfortable and cool on an otherwise ugly day.

What does this mean for me, the consumer? Very little. I pay the same rate regardless of heroic effort. Who cares about power balancing, marginal prices, or what the generators had to do in order to get me the electricity? Turn it up and let it rip!

This is an insane way to run a market. Would you act any differently if you knew that your electricity cost was going to be 4x higher today? Personally, I'd be swimming in a lake.


My Mobile App: BostonBikeLane

A while ago, I asked for your help in developing a smartphone app for cyclists. The 6.083 mobile app development class is now over and I proudly present the completed app. BostonBikeLane reports cars blocking bike lanes to the city so that these areas can be targeted for additional enforcement. This short video explains it all:
Next step: present to the city!


Strategic Consulting in Cleantech

One of the best things about the SDM program is that it gives you some powerful tools, then encourages you to go out and use them. The Systems Leadership and Management (SLaM) Lab pairs teams of students with companies that want a fresh approach to a business challenge. My team worked with local information-driven energy management firm EnerNOC. We came up with some unexpected solutions and everyone benefitted greatly from the engagement. Check my recent SDM blog article on how we provided strategic cleantech consulting:

Throughout the course, the topics included leadership, teamwork, and group dynamics. Team members were able to apply these concepts immediately and improve their critical soft skills. Swope Fleming (SDM'10) summarized the experience well: "I thought it was an excellent project and EnerNOC's support was fantastic. Being able to work with 'live ammo' is something that just cannot be replicated in the classroom alone, and it was an invaluable experience."


Thesis Update: Wind Power and Demand Response in Denmark and Hawaii

Today I signed up 500 MW of demand response customers in Denmark.

Well, simulated customers.

My thesis, you may recall, is on integration of wind resources in Hawaii. Denmark is a pretty good analogue for Hawaii since it has lots of wind and a similar number of thermal generators. There is also a really good stochastic unit commitment model for the nordic countries called WILMAR. I am in the process of modifying WILMAR so that it can simulate Hawaii, but before that I wanted to see how Denmark would behave if it had a lot of responsive demand and 20% wind.

Rather than treating responsive demand as actual curtailment, I model it as another type of generator and leave the demand function unchanged. On the graph above, you see coal/oil/gas powerplants in grey, wind in green, and demand response in red. The x-axis represents about two weeks of operations. To model demand response, I am considering it (to first approximation) to have instantaneous spin-up time, an infinite ramp rate, and no startup/shutdown cost. Its marginal cost is higher than all other generators in the market.

Given these characteristics, I expected DR to be used relatively infrequently when the wind came in less strong than predicted. Not so - the optimization seems to like the flexibility of demand response and dispatches it all the time. I need to look deeper into my assumptions about DR having zero startup cost. (Furthermore, HECO rules promise that DR won't be called for longer than 2 hours at a time. Obviously that still needs to be coded into the constraint set.)

Other interesting observation: that large dark grey band is coal. For all of its world-class wind penetration, the country is still running a lot of dirty coal. You see the cleaner natural gas plants (light grey) kick in mostly when the wind dies down such as on 7/3. Running coal and wind feels to me like eating chocolate cake and diet coke to balance it out. I'd be interested in calculating the carbon-intensity of denmark's dirty/clean hybrid and comparing it to New England's pretty-clean natural gas sector. (Also: 100% wind on the night of 7/12! Go Denmark!)


Patent Hat Trick

After 5 years in the USPTO caverns, the third karl critz patent has been issued. This relates to my time back at the MathWorks as product lead for the Report Generator. Users had been asking for some easy way to package up Simulink and Stateflow models, then send them to colleagues for review. Interviews showed that people were willing to install a special program or plug-in, but preferred not to. I built a tool that would export these models to then-rare Scalable Vector Graphics so that models could be viewed and navigated in any modern web browser. There was even a neat "overview" mode that anticipated Apple's "Expose" feature by a few years. This feature is what made us jump to version 3.0, and it helped sales tidily. My career has progressed onward from this sort of work, but it's still cool to get the recognition.

(It is a bit embarrassing that the lawyers captured the figures using internet explorer, and "provided by compaq" no less.)


Business Logic 2x2

Following my recent realization that for business theory "if it can't be decomposed into a 2x2 matrix, it's not worth knowing", I hereby present my meta 2x2 matrix to describe all 2x2 business matrices:

can it be expressed
as a
2x2 matrix?
yes simplify
it hurts
no go back to
Karl's Business Logic noyes
should it be expressed as a 2x2 matrix?

Note to anyone seeing this through facebook: Mark Zuckerberg hates the <table> tag, so you'll have to click the link to see this on my blog karlcritz.com with original formatting intact.


Marketing Advice for Career Switchers

This week my Harvard Business School class ("Business Marketing") covered the topic of "what is your message when selling stuff?" In the model, there is one particular trap that I have fallen into, as may many of my SDM colleagues. Here is a short warning for those thinking of switching careers.

It seems like there's a truism in business literature that "if it can't be decomposed into a 2x2 matrix, it's not worth knowing." The latest bit of wisdom from Business Marketing concerns itself with how to sell a product based on its benefits. You have to consider your ability to deliver the benefit (low/high) as well as the importance of that benefit to the buyer (low/high). The strategy matrix looks like this:

ability to provide

If you do something well and the customer values it, lead with it. It's your hook. If you don't do something well and the customer doesn't care, ignore it and don't even talk about it. Who cares? If you don't do something well and the customer does care, you are in the realm of objection handling. Deflect and move on; there are dozens of books to help you do this with grace.

Career-switchers get trapped by the lower-right quadrant. We're looking to go into a new area and the bulk of our accomplishments are in a different area. It takes a tremendous amount of strength to resist and NOT play up areas of strength when the customer (a hiring manager) doesn't care. Imagine if someone wants to sell you a car and tells you how great it is at closed-course auto racing. Your thoughts are going to be "I don't want to do auto racing, why are you wasting my time?" and "Am I going to have to pay more for this?" In the best circumstance, the sales message might link that auto racing prowess to things you do care about such as acceleration or cornering. In the worst case, this quadrant will just bother or confuse the buyer.

Prior to SDM, my resume contained a lot of detail about software projects, languages, libraries, and skills. It's been painful to pare this away. I still see myself entering a world of managing software developers or at least working with them to develop specifications. Therefore, my software experience on the resume now focuses much more on leadership and real-world results. How many people did I manage? How did I develop and implement a product vision? None of the jobs I want will care whether or not I can finesse XML XSLT engines to output elegant SVG images, so that line had to go. Keep it relevant to the audience, and be careful what you're selling.

(On a personal note, removing OASIS DocBook skill may have hurt. That pain was nothing compared to removing "Eagle Scout, BSA" from the "awards" section. Darn you, relevance.)

(Image via Wenger, makers of knives both practical and... not.)


SDM Explained: A Transcript

What follows is an organized list of my coursework as part of SDM. The concept of "Systems Design and Management" can be hard to encapsulate in a simple sentence, so this exhaustive list can give you a better feel for how I have been applying myself for the last year and a half. The high-level themes provide the best feel for the SDM approach: leadership, the business of technology, and large-scale systems analysis. My personal area of interest is in the integration of renewable energy, visible both in the explicitly energy-related classes as well as the courses in which I performed substantial project work on energy issues. (Energy-focused courses are noted in bold.)

  • Technology Strategy
    • 15.965 Technology Strategy - Creating and capturing value
    • 15.969 User-Centered Innovation - Listening to lead users for product design
    • ESD.945 SLaM Praxis - How decisions are made, frameworks for competitive analysis
    • ESD.58 Disruptive Technologies - How technologies evolve
      • Project: market evaluation of 3rd generation solar cells
    • ESD.945 SLaM Lab - Applied technology strategy
      • Project: New market penetration for an energy efficiency firm
  • Management
    • 15.381 The Human Side of Technology - leadership theory as applied to tech
    • ESD.930 Leadership - leadership frameworks and personal reflection
    • ESD.38 Enterprise Architecture - an engineering approach to organizational transformation
    • 15.514 Managerial Accounting - how to think about costs and income
    • ESD.763 Supply Chain Management - treating a supply chain as an optimizable design
    • 15.281 Advanced Leadership Communication - running effective teams
    • HBS1929 Business Marketing - how to manage a team for b2b sales
      • Project on how to sell to electrical utilities
  • Energy
    • 15.366 Energy Ventures - Practical lab in how to launch an energy-industry startup
    • ESD.940 Wind Turbine Design - manufacturing, meteorology, and the NREL FAST suite
    • ESD.934 Engineering, Economics, and Regulation of the Power Sector - grid operations
      • Project: future outlook for demand response
    • ESD.865 Modeling Electric Power Systems - numerical optimization with GAMS
    • Thesis - Demand Response for Grid Balance with High Renewable Penetration
  • System Engineering
    • ESD.33 System Engineering - analyzed the Hawaii Clean Energy Initiative as a large-scale systems problem
    • ESD.34 System Architecture - applied architectural principles to the Regional Greenhouse Gas Initiative
    • ESD.36 Project Management - how to bring in a project on time and on budget, using system dynamics to avoid the traps
    • ESD.344 Real Options - mathematical tools for when and how to build flexibility into product design
  • Misc
    • ESD.301 Statistics - Emphasis on hypothesis testing
    • 21F.415 Deutschland im Europ√§ischen Kontext - Deutsche Literatur auf Deutsch
    • 6.083 Mobile Application Development - Wrote 5 deployable Android applications

This portfolio of classes has positioned me for a role in product management, project management, corporate strategy, or engineering management.


Product Camp Boston: Register Today

Registration just opened for Product Camp Boston. The focus is on product management, the difficult and under-appreciated art of figuring out "what to build", not just "how to build it." It's an O'Reilly-style "un-conference", which means that the content will be organized on the spot according to interest. It also has the advantage of being free. I attended the 2009 camp and learned a lot. Let me know if you'll be there.


MIT Energy Conference - Visit the Free Showcase on Friday

I'm on the organizing team for the MIT Energy Conference Showcase. Tickets for the conference itself sold out long ago, but the Friday (4 March) Showcase is free and open to the public. We have put together a great collection of academics and companies doing cutting-edge work in the energy field. I'm particularly excited by some of the unconventional wind and hydro generators we'll have on the floor. Stop by the Westin from 5 to 8pm and say hi.


Dev Project: Boston Bicycling Mobile App

Ruff Ruff

Your smartphone is a sophisticated, connected sensor platform. The City of Boston is rolling out apps to improve quality of life through its "new urban mechanic" initiative. Through apps like Citizens Connect, we can report graffiti, get streetlamps fixed, and (maybe) even automatically detect potholes. I am working on a project with Nigel Jacob, owner of the mechanic project. As a cyclist, I'm prototyping a mobile app that will make life better for bikers in Boston. If it looks good, his department may invest the time to release it as a production-quality system. This would be a great and low-cost way for the Mayor to follow through on his promise to make the city a "cyclist's dream."

Here's my question for you: What do you want? How can we help? What cycling frustrations could be solved with a mobile app? A few ideas to start the conversation:

  • Need: finding bike lanes. Help with route planning.
  • Frustration: cars parked in bike lanes. Submit photos of offenders? Automatically detect swerving into traffic when in a bike lane?
  • Need: automatic dispatch of police/ambulance after an accident. Detect sudden deceleration followed by extended immobility, message 911 with current location.
How can we use technology and the support of the city to make this a great place for cyclists?

Image licensed Creative Commons by abbyladybug.


Strategy and Stupidity

Corporate strategy is often presented as being like a game of chess. (Better yet, "go" or "pente".) You make your plan, seek to outwit the competition, and emerge victorious through superior execution. The dumber your competitor, the more likely you are to win.

I first became aware that this view was incomplete during my startup days. Our product category represented a tiny but disruptive opportunity to a much larger market. Arguing with or belittling competitors wouldn't just be childish, it could shake confidence in the entire nascent industry.

I am taking a B2B marketing class at Harvard this semester and we just saw a video of an executive who stated this wisdom concisely (I paraphrase):

It's terrible to compete against someone stupid, especially if they don't understand their own cost structure. They'll send themselves into bankruptcy and may drag you with them.

It's easy to recall industries poisoned by unethical competitors; energy trading will live under the shadow of Enron for a generation. I can't think offhand of any industries or markets that were destroyed by a stupid competitor, but imagine that it wouldn't be too hard to find one after some research. Are there any on your mind?


Language: Gendered Address

I am taking a German class this term in preparation for potential post-graduation employment through the MIT-Germany program. It's been fun to reacquaint myself with the language and I'm looking forward to it. There's one unusual wrinkle, though: our professor is a visiting scholar from Wellesley (a women's school). Some Germans have a verbal tick in which they insert "meine Damen und Herren" (ladies and gentlemen) into their speech for emphasis or to pause.

Our professor, on the other hand, repeatedly asks questions like, "Wo denken sie, meine Damen, dass Europa beginnt?" Or, "Ladies, where do you think Europe begins?"

Fortunately, the three Herren in the class think that it's more funny than bothersome.

Note: Photo courtesy of Brown University showing an anatomy class for women in 1900.


Product Review: Agloves

I never imagined that on a tech-centric blog I would be doing fashion reviews, but Agloves sit somewhere between "garment" and "tool." When I found myself on a walk manipulating my touchscreen with my nose, I knew something needed to change. Agloves are a liner-thickness glove threaded through with capacitative silver so that every part of the hand registers on the touchscreen. The fit is a bit strange, but they work as advertised.

Because the gloves are thin, I can do precision targeting like typing. Even hitting small countries in Lux DLX is pretty easy. I'm not sure that the full 10-finger response is all that necessary. I'd probably be happy with an index and middle finger. The real advantage to the sewn-in conductivity is that the lack of a conductive pad means that the gloves can be accurate and have a natural feel. Though not as warm as my nice Black Diamond skiing gloves, they seem insulative enough for most urban applications. They're not stylish, but I'll call them attractive enough that they don't embarrass me.

Unfortunately, these gloves are hard to find at brick-and-mortar stores so you may have to iterate to find your right size. I used the company's online sizing tool (which measures only the width of your palm) to order a M/L pair and had to exchange them for an XL. These gloves have very short-cut fingers, so the M/L pair left me with uncomfortable webs of fabric almost behind my knuckles. Even with the XL gloves, I wish the fingers were longer. Their customer service rep was helpful and friendly, replacing my gloves quickly and without fuss. I was a bit bothered that I had to pay for return postage. Given how personal glove sizing is, I would feel more confident buying with a Zappos-style prepaid shipping return label.

If you want to check your mail during the winter, these will get the job done. I wonder if this will start to become an expected feature in most gloves by next year.


Wind Week @ MIT

Wind Week at MIT just wrapped up. I learned a lot in the special-topics class ESD.930, and particularly enjoyed that the visit from Steve Nolet of TPI Composites enabled me to geek out about prepreg laminates and honeycomb cores. The most thought-provoking talks came during the all-day Friday workshop in which we focused on grid integration. Most of the serious work in wind theory these days is on bidding strategies/market design, as well as in figuring out cost allocations for transmission upgrades and standby generation. These all assume that storage is expensive and we have to think about the grid in a different way.

Eric Ingersoll from General Compression raised a different idea: cheap storage can turn wind from an intermittent resource to a baseload generator. He estimates that 200-300 hours of storage will enable any given facility to be fully dispatchable at a combined cost below that of coal. (And that wind + storage is cheaper than wind + peaker plant.) This raises a lot of interesting questions, like "why not use less storage and call yourself a peaker unit" or "why not chuck the wind and do intra-day arbitrage?" But that's not the point of this post; I'm more interested in the general idea of technical breakthroughs making complex hacks unnecessary.

Remember the beginning of the previous decade? In 2000 we knew that the broadcast model for video would be supplanted by something else, but not what. We knew that the TV of the future would be customizable, timeshiftable, and able to access an infinite array of content. The form of the solution was a huge unknown given the tiny pipes available at the time. TiVo connected a hard drive to the cable box. BitTorrent hooked your hard drive to a world of pirated content via your crappy ISDN line. AppleTV 1.0 did the same thing, but wrapped the juicy content in a razorwire shell of DRM. Deep thinkers imagined legal peer-to-peer multipoint distribution systems and even freely shareable files with embedded advertising. We even briefly flirted with the notion that low-quality YouTube clips of teenagers injuring themselves with fire would replace professional media. Broadband is now a reality and we now pretty much just stream studio TV or movies through Hulu, AppleTV 2.0, or Netflix. The infrastructure necessary to engage in broad realtime streaming of video is quite complex with its Akamai nodes and high-speed backbone, but the user model is radically simplified. We have more or less jettisoned those crazy workarounds and now I can watch old episodes of "Arrested Development" on my phone as I walk to the (T).

System limitations necessitate complex hacks that sometimes have shorter lives than the limitations that spawned them. Will cheap storage turn wind integration into a no-brainer before we implement the perfect intermittent wholesale electricity market? I have my doubts in the short run, but wouldn't be surprised if some day in the distant future we look back on Fred Schweppe's revolutionary power system market designs as a transitional hack.