Liberate Your Data

I discovered a cool tool today which might interest my SDM colleagues: Liberate your Google Groups site! Many of us use Google Groups to coordinate our ad-hoc teams. But how do you preserve your team's work for posterity? The liberation tool exports the entire site to a local directory on your machine. Finally - my entire education on a (surprisingly small) flash drive.

I still have most of my course bibles from my undergrad days. It's a real relief to transform these into a more-portable form.


Pitching Fits

The good news: I filtered through almost 400 of the MIT100k Elevator Pitch Competition entries to the top 60, then placed into the top 12. This let me pitch my clean energy idea to 500 people live and I hear that the video will be seen by a few thousand. Unfortunately, I didn't place in the top 3, so I won no money. At the time, it felt like crushing defeat but I'm trying to look on the positive side and call a top-12 placement an accomplishment.

Here's what I think brought my team to the top:

  • Our low-cost/high-effectiveness desalination idea is easy to understand
  • It has a clear, profitable business model
  • The judges rated me as top 5% in "connects with the audience" and "charisma".
I walked out of the opening round feeling good about the pitch and its delivery.

Then came the finals night, which didn't go as well. What I'd do differently next time:

  • I reworked the pitch based on feedback from the judges in the preliminaries. Their feedback was useful, but the pitch became overstuffed with an "all things to all people" problem.
  • I did not connect with the audience. My delivery was uncharacteristically nervous and stilted. I need to reach deeper into that theatrical background and force myself to be ON regardless of whatever else is happening.

We'll be back at the Executive Summary contest in a few weeks. Until then, it's time to get back to interviewing potential customers.


Take Me To Your Leader

After ridiculing CareerLeader, it's time to post about a self-assessment which really works. One of my classes this term is ESD.945 "Systems Leadership and Management". It's a project-based class which looks explicitly at effective teamwork, decision-making, leadership style, and management theory. I have always found it strange that much of being a manager is about leadership, but that business school focuses so much on the nuts and bolts of strategy/economics/accounting. One of our early exercises in this class was to take the "Belbin Team Roles" test. I have to say that I was surprised at the results.

"Know Thyself, Algorithmically": The test prioritizes your teamwork style into a variety of iconic roles. I had always seen myself as a dispassionate, technocratic "coordinator": the one who doesn't care much about the direction of the team as long as everything is humming along efficiently. As it turns out, I'm more of a "shaper", a description which fits over 80% of business school graduates. I have a vision, and dang it if we're not moving in that direction. It's either time that I own up to this tendency, or explicitly walk away from it. Denial helps nobody.

I at least have one trait which is rare among MBA-types. My second-highest score is as an "implementer". Though the name sounds like an individual contributor, this role actually shows strengths in delegation, leading the team towards its goals, and generally making sure that Stuff Gets Done. Implementers are apparently rare in business school, so I do have an edge here.

My team for this class contains another strong "shaper" personality. We're already talking about how to best make sure that we work constructively. This is a conversation which probably wouldn't have happened without the test. Call it an early win.


Electricity Grid: System Balance and Renewable Integration in Hawai'i

The following is reprint of an article I wrote for the Fall 2010 edition of the SDM Pulse. Read it in all its PDFy glory there, or enjoy the web form here. This project inspired what has become my thesis topic, so you'll be seeing much more about Hawai'i here. The project team consists of Kacy Gerst (SDM ’09), Matt Harper, (SDM'10), and myself Karl Critz (SDM ’10).

The state of Hawaii has great incentive to pursue renewable energy projects. The Hawaii Clean Energy Initiative (HCEI) has provided top-down pressure for change by setting targets of 40 percent renewable energy and a 30 percent increase in energy efficiency by 2030. Electricity costs triple the national average, and gas at the pump is 50 percent more expensive than on the mainland. This combination of political and economic drivers encourages Hawaii to test new systems for energy efficiency and sustainability significantly before such systems are explored on a national scale. For this summer’s course in systems engineering, SDM students undertook to find ways to help Hawaii reach its targets in these two areas: increased transportation efficiency and stable renewable electricity production.

Much economic and policy research has already focused on how to structure incentives in order to meet a 40 percent renewable portfolio standard. Team Grid therefore chose to focus on the less-studied systems issues and incentives involved in integrating intermittent sources of energy, such as wind and solar power, into the electrical grid. While some renewable energy sources, such as biomass or biofuel, act like the status quo fossil fuels and can be ramped up or down as needed, others do not. Geothermal energy installations usually have a fixed maximum capacity and have limited ability to respond to demand variation. Worse, wind and solar sources are entirely at the mercy of nature. A grid supplied by these intermittent sources must work harder to meet demand when the wind stops blowing.

The SDM team deployed a rich set of systems engineering tools to address the problem. Characterizing the proposed electricity grid for 2030 exposed the scope of the problem on the minute and hour timescale. A model of the grid from today until 2030 revealed the connections between generation, demand, investment, equipment retirement, transmission, and stabilization. The team also developed a model of stakeholders to put hard economic values on the cost of blackouts, not-in-my-backyard attitudes, habitat destruction, and behavior change. By using experimental design, the team evaluated a set of portfolios for its economic and social costs. This analysis revealed an optimal set of stabilizers for assuring an adequate energy supply with intermittent resources.

The best portfolios focused on simple solutions that use existing infrastructure. It is not, strictly speaking, economical to maintain oil-fired power plants when they will only be used infrequently. However, compared to other storage technologies it is much less expensive (economically and socially) to keep these plants maintained and ready to step in when the sun and wind cannot provide. The team therefore recommended that the Public Utilities Commission guarantee that low- utilization oil plants be compensated by ratepayers.

Unfortunately, such plants are unable to take extra energy when the wind is blowing strong and demand is low (“down-regulate”). To react quickly to unexpected changes and stabilize the grid, the team also recommended the use of chemical energy storage devices such as batteries or fuel cells. Since the storage would supply broad grid benefits, it makes sense that it be controlled by the electric company and not by individual wind/solar developers. The general benefits of storage should also qualify investments for public subsidies similar to the producer tax credit offered for wind and solar developers.

In addition to these two themes, the team also found benefit in (1) dynamic billing policies to shave demand during peak times and (2) streamlined siting for transmission lines and geographically distributed intermittent sources. Each of these policies will create the strong grid Hawaii needs to reduce its fossil fuel imports and assure the continued services upon which its economy depends.

These projects were developed in close collaboration with Michael Duffy at the National Renewable Energy Lab, who provided continuous feedback and guidance to SDM students throughout the course. The teams thank Duffy (who received a master’s from MIT and a PhD from Ohio State University) for his mentorship.


Your Opinion, While Interesting, Is Irrelevant*

With the recent introduction iPhone OS 4.1, Apple is creeping forward in functionality. Much like iOS4, this latest release mostly contains enhancements which I can't use (HDR photography is iPhone4-only) or don't care about (GameCenter/Ping). One welcome and less-publicized feature is full Bluetooth AVRCP support, which means that I can finally use "forward" and "back" controls when the phone pairs with my car.

The iOS platform now represents about 1/3 of Apple's revenue, so it's hard to argue with success. There are dark corners of the user experience on this device which haven't received much attention since launch. As a product manager, these would annoy me like a sore tooth. There is a truism of product development which states "If you're not embarrassed by your v1.0 product, you waited too long to release." All products, no matter how mature, will always have some sort of deficiency. Deciding which ones are important to your customers and which ones just irritate a niche subset is a critical skill.

* Title inspired by Pragmatic Marketing, taking the ego out of product management one framework at a time.


Entrepreneurship: The Flip

I'm at Startup Bootcamp listening to Mick Mountz of Kiva Systems talk about his early experiences with the company. (They make a way-cool automated material handling system which turns warehouses and distribution centers into a robotic ballet.) His step 1 in a startup is (of course) "get a whiteboard and a business card." With a good idea and some customer-funded development, this eventually lets you change the world.

Step zero, though, is something I had never heard before. He calls it "The Flip". It's the moment when you decide "fuck it, I'm doing this." After the flip, he answered his phone "kiva systems". He had no product, no customers, no staff, and hadn't even incorporated. But it became real in his head.

I have heard talks on every detail of entrepreneurship from inception to funding to exit. But this is the first time anyone has talked about something like The Flip. I like the intentionality of this concept. It feels like there needs to be a ceremony for this.


My colleague Rafael told me about the MIT scripts service, which means that I now own karlcritz.mit.edu. It simply redirects to this blog, but it's a cool URL to own. To my other classmates, this takes about 5 minutes to set up. (Though it helps to know a thing or two about MIT's athena unix environment.) Thanks, Rafael!


CareerLeader - This Much We Know

It's hard to provide recruiting support for SDM students. We have 5-7 years of experience and a rather novel degree which combines management with a hard-to-explain corner of engineering. On-campus recruiting and career fairs are targeted at more entry-level positions or (at Sloan) investment banking/management consulting. SDM-appropriate jobs are often not advertised; companies may not even realize that our expertise exists. One service offered by our career services department is the "CareerLeader" assessment. It is supposed to help you sharpen your focus and figure out what kind of job to look for. Having taken the test, I'm not sure it told me anything I didn't already know.

The results are near-tautological. One section asks the test-taker to perform a series of binary rankings: "Would you rather be recognized as brilliant or would you rather make a social contribution with your work?" After a near-comprehensive search of the trade space, we are presented (voila!) with a ranking of our preferences. In another section, we are asked to rank our interest in such areas as "managing technical projects." Indicating high interest in this area means that the summary will tell you "you are interested in managing technical projects!". Again, the insight is stunning. I am not sure what I expected out of this assessment, but I was hoping for some emergent insight into my interests rather than a regurgitation of my answers.

For what it's worth, here's what the test told me:

  • You have a strong interest in Application of Technology. You take a systematic, engineering-like approach to solving problems and understanding systems and processes.
  • You also have a notable level of interest in Theory Development and Conceptual Thinking. You enjoy solving business problems by taking a conceptual "big picture" approach, exploring abstract ideas and the "what ifs" of a business or industry, and considering broad economic and social trends.
This is pretty spot-on, if obvious. Furthermore, the test tells me that careers I should consider are:
  • R&D Management
  • Product Management
  • Management of New Product Development
  • Entrepreneurship
Again, it's just what I told the assessment but at least it's accurate.

Thanks to the hard-working SDM career office for trying on this one. I'd be interested in hearing from you if you extracted value from this assessment. Did it tell you something you didn't already know?


Dirty Power / Shift the Baseline

At an MIT Energy Initiative talk last week, a representative from China"s electric company told us about China's aggressive green energy goals for the next few decades. On one slide, he talked about reducing carbon emissions by 45% per unit of GDP. This sounds great, until you realize that China expects to grow its economy 6x in the same period, so you are looking at a 3x increase in carbon. Putting the two slides together, I felt a bit hoodwinked.

This has been a constant point of contention between the developed and the developing world. Our post-industrial economies hum along efficiently and provide us with a standard of living unimaginable to our grandparents. Meanwhile, 5 billion people are struggling to lift themselves out of a pre-industrial existence. Instituting a nation-by-nation carbon cap freezes the status quo in a manner that probably should be unacceptable to a group of people who largely don't even have indoor plumbing.

In a static, closed system it shouldn't matter much whether you auction off a quantity-limited set of carbon allowances or impose a carbon tax. The magic of price/supply elasticity dictates that the outcome should be the same either way. But that magic only applies within one nation. How do you decide ahead of time the allowable carbon output for each nation? Can you? Should you?

Would you set caps or targets normalized to GDP as implied by the Chinese representative? Could such a system kick an economy while it is down? (A local depression would reduce your relative carbon allowances, making it harder to recover.) Is is verifiable?

There is a simplicity in taxing emissions rather than auctioning quantity limits. A roughly consistent worldwide price on carbon would automatically and dynamically adjust emissions as economies develop and become more or less energy intensive. It is much more transparent and implementable than a single global allowance auction. It would enable developing nations to grow and increase their absolute and relative output while still recognizing the cost of that growth.

Nobody wants to mire developing nations in poverty. Nobody wants them to have a free ride either. We need a regulatory framework which recognizes that 1/3 of the world is growing faster than laws or treaties can adapt.


SDM Summer 2010 review

Another semester, another chunk of knowledge. Summer term is a bit accelerated at MIT since it tries to pack a full semester of content into fewer weeks. Classes meet for more hours each week and the workload per unit time is higher. Fortunately, the institute is air conditioned.

ESD.763 covered supply chain logistics from a very high level. I didn't walk out of this class feeling like I could design and run a supply chain on my own, but I now feel that I have a solid grounding in the issues and the fundamentals. We started with some good mathematical grounding in Markov chains and queue theory. Then we covered a rich set of cases to see how these theories are applied in the real world. Learning about 7-11 Japan was mind-blowing. Hourly resupply with shifting stock throughout the day? Insane. This class only ran for the first half of the semester; having its workload disappear just as other classes ramped up was scheduling genius. I appreciated that Professor Martinez-de-Albeniz would challenge poorly-thought-out answers with harsh and well-deserved reality.

15.514 Financial and Managerial Accounting should have been painful. Who wants to spend three evenings per week learning about accounting in the summer? It turned out to be much more engaging than I thought it would be. I viewed the class as "defense against the dark arts." I never plan to do any accounting myself, but I feel that I should now be able to spot and understand a severe irregularity if someone tries to sneak it by me. Professor Scott Keating was committed to both our understanding and keeping the material as engaging as it realistically can be.

ESD.33 Systems Engineering reminded me of my employee training on the "Design V" back at Ford. This was mostly familiar material, and it suffers a bit by being so loose and heuristic-based. It's necessary material but I don't see a systematic way to teach it. This class, however, was responsible for my favorite project of the summer: an examination of the electrical grid in Hawaii. (Separate article on this to come.)

ESD.945 SLaM Lab Praxis was the first formal "leadership" training we have yet had. In this class we covered decision making, setting strategic directions, how to disagree productively, and how to respond to changes in the external environment. This was good material and Professor Michael Davies delivers it with passion. There is a good chance that I will be taking the followup class this Fall.

And now, three weeks off for some backpacking/kayaking/thesis-writing.


Wave Goodbye

Google has stabilized Wave. The occasional miss should not be a surprise in a company with a freewheeling, beta-oriented culture. There is something to be said for the argument that Wave was poorly communicated. If you have a new product and can't describe what it is in a paragraph of text, you have a problem. Who is it for? What do you do with it? How does this make your life better?

My theory is that Wave failed because of its invitation strategy. Wave is a communications platform. If you can't use it to communicate with anyone, it's useless. In school, we frequently form fast ad-hoc groups for projects and problem sets. Wave would have been perfect perfect for this since email is too unstructured and Google Groups is too heavyweight. I tried early on to sell a few of my teams on Wave, but quickly gave up when we ran into invitation/registration problems. When you only have a week to get your team moving, tool overhead is something you can't afford.

Google's Buzz has been maligned for lacking any meaningful functionality, but it did the right thing in becoming available for everyone, everywhere, immediately. I appreciate Google's caution in wanting to ramp Wave in a controlled fashion, but a slow rollout can be death for a social product.


Lines, Spirals, and Product Development

A talk this week on product development included this diagram about the "product requirements death spiral". As befits NASA, it casts iteration in development as a black hole which can only be escaped by finalizing a design and achieving escape velocity. A different part of the same talk covered "Death, Taxes, and Change Requests." When working in heavy iron industries like aerospace or automotive, I can understand a certain weariness about design changes.

The telling comparison is to look at the agile software iterative development spiral. It assumes and even celebrates changes, viewing development as an experimental process. Lessons learned from one prototype are applied to the next version. Changes aren't the exception, they are part of the process.

Once upon a time, I got frustrated when a broad-consensus group design decision was reversed later in development. It seemed inefficient to backtrack. Eventually, I learned that paper prototyping and whiteboard sketching can only take you so far. Stakeholders will always change their minds. In fact, they should. We learn so much by seeing an implementation in the flesh that the product can become markedly better if we listen to those lessons.

Lately I have been wondering how to port this more nimble and experimental software model back into the physical world. I enjoy working on software precisely because of this level of freedom. But there is a lot of appeal to building something out of atoms instead of bits. I wonder if someday 3D printing and rapid manufacturing will enable true iterative development on heavy iron.

Disrupting Photons

Technology-watchers have been predicting that cell phone cameras will replace single-function cameras since the first featurephones were released. The serious photographers who write on the subject have now embraced the digital SLR over chemical film, but still laugh at the idea that the tiny optics of a phonecam could ever subsume a dedicated device. Today, I saw evidence that the laughing is over and that phone manufacturers have moved on to the next phase in the tech transfer attitude chain. ("first they mock you, then they fight you, then you win")

There is a billboard just outside Logan airport which says "If it has a ringtone, it's not a camera." This is obviously a defensive move from a threatened incumbent. I am writing this from MIT's Killian Court where hundreds of tourists pour forth from buses each day to photograph each other in front of the great dome. An informal survey shows that about half are using cellphone cameras. Personally, I take more pictures with my blurry iPhone 3GS camera than with my fancy Canon s90 or my rugged Pentax w80. The iPhone 4's camera is even better and will probably represent even more of my picture-taking when I purchase its successor. The idea of a cell camera replacing a single-purpose device is already here for the majority of consumers and is just around the corner for everybody but the most serious artiste.

My tech strategy classes last semester covered many such turning points. We have seen the signs and they are all here. I am imagining the New England lake-ice industry with a billboard in 1920 saying "If it came from a machine, it isn't ice." Sailboat makers could claim "If it has propellers, it isn't a ship." Gas lighting manufacturers might try, "If there is no flame, there is no illumination." This campaign comes off just as desperate, and just as doomed.


Semester 1 in the rearview mirror

My first semester is now done. Some thoughts on my classes:
  • Disruptive Technologies: We learned a lot about the Clay Christensen innovation model during IAP. It was presented as a well-packaged, clear, consistent model for understanding the dynamics of the marketplace. This class is an extended series of footnotes, caveats, and counterexamples which show that the simple model isn't the end of the discussion. Group project: understanding the impact of low-cost 3rd generation PV solar technologies. We predicted that Balance of System costs will keep 3G solar out of the utility-scale market indefinitely, but that opportunities posed by BIPV (Building Integrated Photovoltaic) to lower BoS costs will create a viable market once the technology's durability issues are resolved.
  • Engineering, Economics, and Regulation of the Electricity Grid. My favorite class this term. A detailed, real-world look at exactly how the magic of always-available power actually works. I had no idea that the system was this complicated and how the incentive structures really work in the industry. Fascinating to learn what "deregulation" really means.
  • Technology Strategy: Heavy overlap with Disruptive Technologies. (same instructor) It was cool to read some seminal books of the field and then have the opportunity to interview the authors at length in class.
  • Real Options: "How to introduce a stochastic distribution function to your business plan". Really useful content burdened by an unwieldy class organization. Downright painful use of inappropriate tools to the problem. Unpleasant for structural issues, but the material is compelling enough to save the class.
  • User-Centered Innovation - discused earlier
  • Thesis Seminar - I may have a thesis topic soon. Yay.
  • Enterprise Architecture - A not-yet-rigorous approach to enterprise transformation. I had hoped for an engineering approach to management, which is not really what this class is about. Good to know, though, especially if I ever need to roll out an IT system in a large company.
Of course, the most valuable stuff at MIT comes outside of classes. I volunteered with the Clean Energy Prize, which deserves a post all of its own. Stay tuned.


A Dynamic Pricing Thought Experiment

Imagine that all residential electricity customers experience real-time pricing. We all have smart meters and smart appliances which enable customers to register a spot price beyond which the user is willing to automatically curtail use. Imagine that a popular air conditioner manufacturer ships its unit with a default curtailment price of $.25/kWh. What is your ideal bid?

You certainly don't want to leave it set at the user default. Your bid will be lost amid the others. If you value your comfort, a bid of $.26 will clear out the vast mass of people who just leave their appliance set at the default while still preserving the lowest possible bid. I'm not sure what a $.24 bid says about you, though. Maybe that you're cheap and want to capitalize on savings before everyone else?

ERCOT (in Texas) already limits the pool of demand resources because they can react so fast to an event call that the system goes into overvoltage. I wonder if device manufacturers will be required to ship their smart appliances with a randomized default bid to prevent a sudden shutoff of half the air conditioners at a substation.

The world of real-time pricing is going to be a fascinating place in ways the academics haven't even begun to consider.


What I Like About Grad School

As the usual end-of-term crunch sets in and forces me inside on perfectly nice spring days, it is important to remember why I am doing this and what I like about it. Basically, it comes down to "no bullshit". The undergrad experience (especially at MIT) often relies on frequent testing in order to filter out the ones who don't belong in the program. Grad school assumes that you want to be there and treats you like an adult.

few textbooks I have had a few professors require expensive textbooks, often of their own books. But most of our reading is in the form of journal articles or other papers. Throw in a few eBooks and I like that few trees had to die for my degree. Most of the books I have purchased so far have been inexpensive business mass-market paperbacks, so that annoying nickeled-and-dimed feeling doesn't set in at the beginning of every semester.

no tests, no finals Grading has so far been based entirely upon papers, presentations, projects, and participation. With no final exams at all, I can relax during exam week and actually get some thesis work done. I'm learning quite a bit on my project work, whereas tests tend to be about demonstrating mastery instead of learning anything new.

educators as equals I can't break the habit of calling my professors "Professor Lastname", even though several have told me that it's ok to be on a first name basis. The honorifics are there because I feel they deserve it. But this time around I see my professors as (more-knowledgeable) peers, rather than unimpeachable wise men. This means that I'm more likely to challenge them, wouldn't feel as awkward having a beer with them, and understand that their role as lecturer/teacher is actually a rather small part of their life.

Education for education's sake My mental model failure as an undergrad was to consider the program as a series of hoops to be jumped through in order to get a degree. This time around, I figure that nobody is going to care about what degree I have. My future prospects are almost entirely a function of what I learn, who I meet, and how I acquit myself in the process. Classes and grades matter, but they are not arbitrary obstacles. (I particularly like that the SDM program director is pretty open to waiving class requirements if you can make a good argument.)

There's more, but final projects call. Let's see if I'm still this chipper in 3 weeks.


User-Centered Doneness

Spring break* is over, we're halfway done with the semester, and my time in SDM is already 1/8 completed. If I'm 8x more knowledgeable next May, I'll be unstoppable. (Sadly, diminishing returns means that this is unlikely.)

I'm already done with one class, "User-Centered Design in the Internet Age" taught by Eric von Hippel. It consists mostly of ideas you could get by reading Slashdot for a few years, but backed by academic rigor instead of uninformed speculation. It's good to see that open platforms, user-driven innovation, and free sharing do have a place in a world of hard-headed analysis, ROI, and NPV. Open source: it's not just for ideologs anymore. (I liked that von hippel believes his own talk: he released his book creative commons!)

It was good to see corporate/military types who thought that open source is for unwashed hippie hackers realize that the spirit of open innovation could work for them. But if I had to improve the content of the class, I would suggest that we look more at where the user-centered approach fails and why. These techniques are great in some places, disastrous in others, and hard to quantify elsewhere. How deep should an organization go?


MIT Energy Conference 2010

This weekend is the MIT Energy Conference. There are more interesting lectures than I can possibly attend, and that's even including skipping one of my favorite classes. (Sorry, professor von hippel.) I won't summarize every talk, but a few deserve mention:

The provocatively titled "Is green energy the next bubble?" panel talk provided some good fodder for thought. The consensus was that since the market for green energy is mostly provided by government policy, it is unlikely to overheat and bubble. But there was some good criticism of ethanol, which one panelist artfully described as a "bridge to nowhere." If incentives create powerful constituencies for other uneconomic policies, could we get into a similar bizzaroland? Another panelist, though, mounted an interesting defense of ethanol. Her argument was that the first generation food-based feedstocks created a market into which more-reasonable second-generation feedstocks (like cellulosic) could enter. I'm still not holding my breath.

The panel on electric vehicles asked a particularly incisive question: "If EVs are the solution, what is the problem?" Tellingly, the justifications were all over the place. One speaker mentioned "carbon", which is great (assuming you're not powered from a coal plant) but I don't see the incentive mechanism which will make drivers WANT an electric car. Another mentioned "energy security", which at least has a price mechanism (oil futures internalize uncertainty). The CEO of Fisker hit closer to the truth - driving enjoyment and ability to travel in emission-free european city hubs. Like all new technologies, EVs will continue to disappoint when compared to to internal combustion in the traditional metrics of range, weight, and cost. Without a price on carbon, the drivers which will bring us EVs will be noise, avoiding gas stations, acceleration, and other new metrics.

A few annoyances:
  • There were a disturbing number of speakers whose presentation consisted of reading numbers verbatim from a paper statement. Boring.
  • The panel discussion on solar had a strange format, with each speaker beginning with a long statement addressing the same 3 issues. I would have preferred that each speaker address a single question in series so as to promote more interplay of ideas.
This was a good conference, very well-run. Perhaps I'll help organize it next year.


Industrial Relations

Because a crushing courseload is not enough, I have committed myself to chair the "industrial engagement" team of SDM's Industrial Relations Committee. Today I finally found out what I have gotten myself into and it seems better-defined than initially feared. The outreach chair figures out ways for companies to partner with SDM. Some companies sponsor students to attend SDM with full salary, which is a pretty maximal level of engagement. But there are plenty of other ways for a company to benefit from a relationship with the program. Here are a few we have identified:
  • Be speaker. SDM sponsors regular events with real-world speakers reflecting on their experiences. Share your wisdom and make contact with talented students.
  • Be a class project. Most of our class work involves longform team projects in product development, market research, and business strategy. Students like solving real-world problems; put a team to work for you.
  • Be a judge. Many of our projects and presentations are judged by industry experts. See what we have done and evaluate it.
  • Be a visit. Schedule a tour of your facility and show off what you are most proud of. Ask us tough questions about what we see.
  • Hire an SDM. Our business skills, engineering understanding, and industrial experience make us a strong fit for any organization.
  • Sponsor a thesis or a research project. Develop a relationship with an SDM student and solve your toughest problems.
  • Request a research summary. With 20 years of SDM thesis work, chances are that there is already a strong body of research in your area. We can consolidate previous work in a compact, targeted format.
My next trick: find companies who are interested in these relationships. There is much to look forward to in the next few months.


10 Years of Educational Change and Stagnation

I wondered how being a student would be different now from 10 years ago. The basics lecturing are still the same. Good instructors are still good instructors. My approach to education has definitely changed, but that's a topic for a different post. I was more concerned about technology. Would pervasive WiFi kill my ability to concentrate in class? Would I check my RSS feeds every time a lecture got boring?

By and large, not as much as I had feared. I do slip occasionally, but my method for handling a boring lecture is about the same now as it was back then: parallel process by doing problem sets in class. I can get more done in class now because the range of projects I can take on from the chair is larger. But I have so much to do that I can't really imagine firing up google reader just because I can't afford the time hit.

I have developed one technique to ensure focus in class, though. When I need to edit text (either by writing a paper at home or taking notes in class), I use OmmWriter. It's an idiosyncratic little text editor with almost no features. There isn't even search and replace. What it does do very well is to present an unbroken full-screen editing surface with no interface elements. You can't really do anything BUT write. If you try to alt-tab over to Firefox to check email the entire text editor disappears, taking with it any pretense that one is "multitasking". This may be the best program ever for dealing with the pervasive distractions which kill creativity and focus.

When I want to crank the text-editing focus up to 11, I'll disable WiFi. This is hard, but it does extend battery life.

Some professors have a different approach to the problem, which is to ban all laptop use during class. I can see where they are coming from. They want us paying attention and not updating our Twitter status. It's understandable, but also completely barbaric. I type 90WPM and can do it heads-up without even thinking. My written word requires looking at the page, is 4x slower, near-illegible, not searchable, and rather difficult to back up. Which would you rather use for taking notes in class? I find it puzzling that I am forcibly relegated to using pre-Gutenberg technology while simultaneously learning about innovation.

So there it is: less has changed than I imagined.


IAP: Done. Spring: bring it on.

Ask any SDM about January and they'll tell you that it's intense.

Week 1 is dominated by the LEGO design challenge, with only a small amount of homework distracting from robot-building. In the beginning I left school around 11:30 in order to make the last (T). As the competition drew closer, I stayed until the wee hours and then gave up and brought in my sleeping bag for an overnighter. Following the competition, I was feeling pretty worn-out and slept for 14 hours straight.

Weeks 2 and 3 are shared between design challenge 2 and a bevy of homework assignments. Classes and lectures were scheduled from 8:30am until 9:30pm. The hardest part was scheduling all of my teams for our group work assignments. Usually the only time everyone could meet was during meals or after our late-night lectures. With some serious frontloading of problem sets onto the weekends, I managed to avoid staying up too late and usually made it home around 11.

Somewhere in here I organized a dim sum run to introduce my colleagues to chinatown. Consider that a small victory that a dozen SDMs were able to take a morning off right before the DC2 presentations were due.

By week 4, we were done with most of our classes and only working on design challenge 3. Again, heavy frontloading saved me from doing much problem set work here. I was able to relax, enjoy the scavenger hunt, and even make dinner for my wife.

3 weeks of nonstop activity is not too bad. I have a theory that a human being can endure just about anything for 6 weeks. This first drink from the firehose in a decade reminded me how to be a MIT student in a way which will make spring term seem like cake.

I hope.


Among My People

One of my first sources of worry about the SDM program was "Opportunity Sets". Not their existence, but their name. The kind of individual who would name a "problem set" an "opportunity set" seemed to me infected by the worst elements of contemporary managerese doublespeak. When I took a break from the corporate world for the academic, I thought I was leaving "learnings" and "incent" behind. Did this mean that I would be surrounded by suit-wearing phonies who actually believed in their golf-themed Successories posters?

Turns out that it's not a problem. My colleagues are My People, engineers and scientists with a leadership-oriented bent. They are honestly interested in solving real technical problems. They have accomplishments from aerospace, civil engineering, IT, consumer electronics, the military, and even the food industry. They have all seen Star Wars. (Except one. We're working on him.)

That said, we all learn the most from people who are different from us. It's comfortable to be surrounded by people Like Me, but I may have to stretch myself a bit to try new things. Does this mean that I may end up in the Sales Club? Stay tuned.

But no Successories posters, please.


Design Challenge - LEGO vs Sleep

I had an ambitious idea that I would be posting regular interpretations of my MIT SDM experience on this blog.


The january program is intense enough that there isn't enough time for sleep, much less regular writing. Heck, I'm writing this during a lecture right now.

Week 1 was dominated by Design Challenge 1, the goal of which was to build a robot which would perform a variety of different tasks. I have been burned by similar challenges before (2.70!) and learned then that the most important thing is to test, test, test. Our team agreed to physical and code freeze by thursday night, then spend all Friday testing. We also had an idealistic vision that we would not work too late any given night. We didn't meet either of these goals totally, but we did manage to spend most of Friday testing. Unfortunately, we tested in the wrong lighting conditions. At the competition, our robot kept seeing its own shadow and confusing it with the lines on the field. Because we tested in a dimly-lit room without shadows, we thought that our robot was 80-95% reliable depending on the event.

Instead, we ended up failing miserably. We scored NEGATIVE points on 3 of the 5 challenges and came in last place. Given how much time we sunk into the challenge and the strength of our test runs, this was a crushing result. I went home Saturday night sleepdep'd and numb.

Once again, the lesson is test, test, test. Except that if your tests don't reflect field conditions accurately, you're not really testing.