Research

Best Poster Award at BuildSys 2011

2011/11/30

en Research

Our poster titled “Appliance Classification and Energy Management Using Multi-Modal Sensing ” won the best poster award at the 3rd ACM Workshop on Embedded Systems for Energy-Efficiency in Buildings (BuidSys 2011) .

The work that we were presenting was based on some preliminary results that we have for the following idea: using data from a variety of sensors in a home (motion, light intensity, plug-level meters, etc.); can we automatically categorize each appliance in the house by finding clusters of correlation between their usage patterns and the activity of the residents?

We found that using a simple statistical correlation as a metric to cluster different signals together, we could determine whether an appliance was “active” (i.e., its usage is highly-correlated with a person being close to it), “passive” (only correlated when the appliance starts) or “background” (not correlated with motion data).

This information is useful for energy management purposes (e.g., reducing energy waste, load shedding and scheduling), and our preliminary results show that it may be possible to determine these categories in an unsupervised fashion.

{ 0 comentarios }

Evaluating Energy Disaggregation Approaches

2011/11/29

en Research

Face Detection via OpenCV

Face Detection via OpenCV. The image belongs to Wikimedia Commons.


Imagine that you are developing a new face recognition algorithm and someone asks you how good you think it is. Let’s say that you are working for Apple and you are developing the latest version of iPhoto, and it is your boss who asks this question. Which of the following answers sounds more appropriate?

You say that your algorithm can, on average:

  • (a) Detect A% of all faces in any given picture.
  • (b) Identify B% of al the faces that it recognizes on a given picture.
  • (c) Detect and identify C% of all faces that are present in a picture.

Regardless of which answer you select from the above, do you think this answer alone would satisfy your boss? Probably not.

Many questions still remain. For instance, there are questions about your evaluation references:

[click to continue…]

{ 0 comentarios }

Research Projects

2011/11/27

en Research

During this experiment of daily writing I have not yet discussed my current research projects. The section on this website that is dedicated to that topic is in need of a serious update, so I will try taking advantage of this promise to write every day and will attempt to share more details about what my students are working on these days.

Particularly, I plan to expand on the following projects:

  • Evaluating Non-Intrusive Load Monitoring Algorithms: how to objectively evaluate different solutions to the electricity disaggregation problem.
  • Context-Aware Smart Appliances: self-managing controllers for allowing small electrical loads to provide grid-level services.
  • Large-Scale Infrastructure Monitoring: understanding the information and technological requirements for the next generation of instrumentation systems for our aging infrastructure. In particular, I may discuss the case of dams and levees
  • Comparing Models of Buildings and the Reality: formalizing the process and metrics used for understanding differences between the models that we use to design, construct and predict the behavior of buildings, and how they actually operate.

{ 0 comentarios }

Google-assisted Narcissism

2011/11/18

en Research

Every now and then I perform a Google search with my name: “Mario Berges”, to see what comes out. I rarely find things that I did not already know about but some times, and just some times, there are some interesting results. For instance, there was the time I discovered that someone with my name had been found dead in Uruguay.

Today, not knowing exactly what to write about here, I again googled my name and found a flyer promoting a book for which I had written a chapter last year. The book is apparently coming out in December this year, and reading my name in the list of “experts and pioneers” made me laugh. Click on the image to check out the flyer:

Embedded Commissioning of Building Systems

Embedded Commissioning of Building Systems

{ 0 comentarios }

On Teaching Interdisciplinary Classes

2011/11/16

en Research

As some of you know, I am teaching a course on Energy Demand and Utilization the second half of this semester. This course is part of the core courses for an interdisciplinary masters program in Energy, Science, Technology and Policy, housed by the school of engineering but not in one specific department.

The students who join this masters program come from a wide variety of technical backgrounds, and they have different levels of experience, and most importantly very different interests and objectives. This makes it very challenging to ensure that each lecture satisfies the expectations of the majority of the students in the class. To make things more complicated, as you may have noticed, the title of the course, Energy Demand and Utilization, covers an extremely wide variety of topics and it is very hard to treat specific topics in much detail.

And that, right there, is what I have come to realize (again) this year when teaching the course.

Here is what we have covered so far:

  • History of energy demand, starting 4.7 x 10^9 years ago.
  • An analysis of the current energy demand: by economic sector, country, etc.
  • Energy Demand of Human Shelters, including heat transfer in buildings, and building technologies.
  • Tools for Analyzing the Demand
  • Smart Grids and an Introduction to AC Power
  • Energy Demand of Transportation and Industry

Those were six classes, two hours each. You can imagine how fast-paced it may have seemed, or how superficial was the coverage of some of the topics.

During the first day of class, I asked each student to tell me what they wanted to get out of this course. The responses were varied, but there seemed to be a larger number of students interested in electricity and power systems. This is what I am trying to devote more time on.

It is impossible to tailor each class to everybody’s expectations, so I think the best strategy will be to try to include topics that are of interest to the majority, and that whatever topic I choose to include I can either (a) cover it in detail during class, or (b) provide pointers to good quality references that can be used to study it.

That is the plan. Let’s see how it goes from here on.

P.S. By the way, this daily writing exercise is definitely helping out. I feel less self-conscious when writing and am slowly getting rid of that fear of writing something useless which typically is what prevents me from doing it more often.

{ 0 comentarios }

The Myriad of Purposes of Blogging

2011/11/14

en Research

Going through my e-mail today, I realized that I needed a place to write down some of the information I was processing regarding achievements that may become useful when preparing annual reports or updating my curriculum vitae. For instance, I am now an associate editor for the Journal of Computing in Civil Engineering. Where should I write it down?

Options:

  • Directly update my curriculum vitae, and hope that I remember to update my website.
  • Tweet it, and then remember to go through all tweets at the end of the month.
  • Write a post on the blog, entirely about this.
  • Keep it in my e-mail archive with a special annotation.
  • Write it down on a piece of paper or a plain-old text file.

Without thinking too much about it (I literally just started writing this as soon as I thought about it), I think that what I need is a “News” section on this website, which is automatically fed by posts from the blog on a specialized category (or tag). If I can figure a way to remove those posts from the general RSS feed, then that’s what I will probably end up doing.

Then it’s a matter of checking that page every time I want to update my CV, etc.

I know… it’s a complicated answer to a very simple problem. But I love structure and organization.

Thoughts?

{ 0 comentarios }

Tools for Estimating the Energy Demand

2011/11/08

en Research

And the day begins. It’s 5:40am, and my inbox has 43 e-mails. I use the characteristic silence of the hours before dawn to prepare myself for the day. This time, I need to work on the the Energy Demand and Utilization class that I teach on the second half of the Fall semester. Today’s lecture will try to introduce a few tools that can help us quantify the energy demand of different processes. The hardest part about this lecture is defining clear learning outcomes that aren’t too specialized or too broad.

Here is what I plan to cover this time around:

  • We will begin the class by recreating, step by step, the Tom Murphy’s argument on a recent article titled “An exercise in galactic-scale energy“. We have already covered the past and present state of our global energy demand, so by allowing the class to briefly think about our future under different growth scenarios, I will hopefully set the context for understanding: (a) the limits of a sustained growth assumption, and (b) the importance of energy accounting. I encourage you to read this article if you haven’t already.
  • From there we will take a slight detour and cover the basics of calculating energy losses in buildings due to heat conduction and ventilation. I will use David McKay’s book “Sustainable Energy Without the Hot Air” as the basis for this part of the lecture. The main objective here is to familiarize the students with the heat loss formulas and with the three major ways in which the energy demand of heating and cooling can be reduced (reducing the temperature demand, reducing the “leakiness” of the building, and improving the coefficient of performance).
  • Lastly, I will very briefly introduce a few relevant tools that can help in the analysis of the energy demand of a system or process. Each one of these tools can be the topic of a full course, so my intention is primarily to define how and when they can be useful. They are: life cycle analysis, engineering economics, system dynamics (specifically casual loop diagrams), linear programming and multi-criteria optimization; and regression analysis.

Let’s see how it goes.

{ 0 comentarios }

Becoming an Expert?

agosto 17, 2011

I have been involved in the academic world pretty much my whole life. I started school somewhere between when I was 12 and 18 months old, and have only been outside of an educational institution for 1 year since then (right after obtaining my bachelor’s degree). Up until I finished the doctoral studies, every step [...]

Read the full article →

Data Acquisition and Instrumentation for Civil Engineers

agosto 12, 2011

The summer is almost over (at least the one in the academic calendar), which means that the corridors of all buildings of campus will soon be teeming with students walking back and forth, no more silence and peace. This also means that I will, again, start teaching. This Fall I will be teaching two mini-courses, [...]

Read the full article →

Smart Meters and Privacy

junio 1, 2011

The California Public Utilities Commission (PUC) recently released a proposal recommending better privacy regulation for smart meter data. Their recommendations were based largely on a report prepared by lawyers from the Electronic Frontiers Foundation and the Center for Democracy and Technology. Let me start by saying that I do not consider myself to be a [...]

Read the full article →

← Entradas Anteriores