Tag Archives: Engineering

It’s Getting Warmer – Morning Data Points

In my previous post I described how I got data from my weather station into MATLAB to help me answer my primary questions:

  • When to open the windows in the evening, to let in cool air?
  • When to close the windows again in the morning?

The data I have available covers 10 days, including a late July heat wave. At first I tried to understand when the lowest outside temperature was reached and when it increased again by more than 0.5 °C and 1 °C compared to the minimum.

DayMin Temperature (°C)TimeTime +0.5°C Time +1°C
18 Jul 201916.305:54 06:5406:54
19 Jul 201917.005:56 06:5207:12
20 Jul 201920.405.56 07:0108:07
21 Jul 201920.705:45 07:0007:56
22 Jul 201919.905:56 06:2707:22
23 Jul 201920.406:14 07:0508:21
24 Jul 201921.405:53 06:4807:33
25 Jul 201922.106:16 06:5107:17
26 Jul 201922.905:42 06:1707:13
27 Jul 201923.105:50 06:3607:36

If I had more data I’d plot a histogram of the times to decide when to close the window in the morning. For now, I can only come up with a “manual estimate” based on looking at the table above. I conclude: During summer I should close the window by 7:30 in order to optimally protect the cool indoor climate.

It’s Getting Warmer – Using MATLAB to Visualize Netatmo Data

Global warming is a thing. There’s no denying that temperatures are rising.

In my quest to learn to live with it, I got a Netatmo weather station that measures outside and inside temperature levels (among other things) in 5 minutes intervals.

My goal is to keep the temperature in my apartment as low as possible without installing AC, so my primary questions are:

  • When to open the windows in the evening, to let in cool air?
  • When to close the windows again?

While the Netatmo website provides a coarse overview of the temperature levels, it’s not particularly helpful in answering these questions because …

  1. the weekly view shows data points in steps of 3 hours
  2. data from different sensors can not be merged into one graph to understand dependencies (e.g., outdoor vs. indoor temperature).

Here’s a plot I generated by exporting the raw data from Netatmo and some MATLAB coding.

Temperature over the course of 10 days.

With all the data available in MATLAB I can now go ahead with extracting some critical data points, such as when the outside temperature reaches its minimum or when it starts increasing again. More on that later.

PhD Thesis Step 2: Collecting ECTS

Finally I figured out which curriculum applies – they have changed it 5 times since I’ve started my PhD. Lucky as I am, the classes and ECTS that I have collected so far can still be used today. Some bureaucratic hurdles later, the university online system now says that I have completed all my ECTS duties. Yay!

image001We may now actually buy the champagne. I’ll keep you posted.

 

PhD Thesis Step 1: Informal Submission

The road to my PhD is a bumpy one. But finally I think I am at the point where it becomes realistic for me to close this nasty open loop within the next 3 months.

The first step (that I am willing to blog about) was the informal submission of my thesis today. I submitted it both to my supervisor and the second, external examiner from the University of Ulm.

phd-coverWe may want to think about shopping for champagne. I’ll keep you posted.

Listen to Fred Hersch

You have to watch this youtube video:

To me Fred Hersch’s version of “Both Sides Now” is profoundly beautiful and touching. It clearly stands out compared to the millions of stimuli that I get on a day to day basis. When I say stimuli I mean everything ranging from music, photography, writing, management presentations, engineering solutions, mathematical equations to coffee table discussions. In my quest to understanding “Life, the Universe & Everything” I can’t resist to ask: “Why?”.

To tell you the truth, I don’t think I have a way to test my answers (hypotheses) in a “true-or-false” manner. However my intuition at the moment tells me this:

  • Technique. You have to be in control of your fingers to be one of the top piano players in the world. Interestingly I don’t think you have to be in total control, as in all points in the four dimensional “what finger, what key, what intensity, what time” space. You just have to cover a big-enough, interesting-enough subspace, which is a challenge already.
  • References. “Both Sides Now” by Joni Mitchell is considered a master piece in its own right. Not only has it proven to capture people’s minds, but for those who know it already, it will most likely trigger positive emotions instantaneously. Similarly Fred Hersch’s interpretation is firmly rooted in the tradition of Jazz, which provides another huge set of references.
  • Story Telling. From the first note to the last Fred Hersch takes you on a journey that is logical from one note/one phrase to the next and that is holistically consistent in a way that it captures your imagination throughout the piece.
  • Personality. Both what he plays and how he plays it screams “Fred Hersch”. Even if you would transcribe his music and have it played by some random, classically trained pianist, it would probably lack some important characteristics.

For lack of scientific methods, I can’t be sure that this list is correct or exhaustive. But what I can say is that I find all of these qualities in Fred Hersch’s “Both Sides Now” and that I should certainly make sure they are in the random coffee table discussion to make it just as beautiful and touching.

Spotlight: Public-Key Cryptography

Public-Key Cryptography is a truly magical thing and I often wonder why we don’t use it more often. We go on and complain about how it is too easy for government agencies to look into our digital lives, while we do have all the tools to prevent it. Only we don’t use them and we don’t seem to ask facebook, gmail and so on to use them – not that they would. The funny part is that it would not require a lot of effort, because computers could automatically do all of that stuff in the background. All that would be needed is that we were able to securely store one, long password (the private key).

In a nutshell public-key cryptography is a mathematical concept that allows us to encrypt a message that is intended to be read by just one person and to sign something to guarantee that it really came from you.

Let’s take a closer look: public-key cryptography is based on a pair of keys, lets call them #1 and #2. When one uses key #1 to encrypt a message one can only decrypt it again using key #2. The same is true for a message encrypted with key #2 – it can only be decrypted using key #1. In other words the two keys are kind of reversible operations that work in both directions. For instance

  • Plain text -> Apply Key #1 -> Cypher Text -> Apply Key #2 -> Original Plain Text
  • Plain text -> Apply Key #2 -> Cypher Text -> Apply Key #1 -> Original Plain Text

but

  • Plain text -> Apply Key #2 -> Cypher Text -> Apply Key #2 -> Nonsense
  • Plain text -> Apply Key #2 -> Cypher Text -> Try something -> Nonsense

In addition to that it is important to know that it is (almost) impossible to find the keys from the Plain text and/or the Cypher Text.

Usually a person or entity (lets call this entity Alice) creates a pair of keys, where one is made public (the public key) and where the other key is kept secret. There are two typical use cases:

  1. Anyone can encrypt a message using the public key and send it to Alice. Only Alice and nobody else (e.g., someone who intercepts the cypher text) will be able to read the plain text when only Alice knows her secret key. – This use case is also called “Public-key encryption”.
  2. Alice can encrypt a message using her private key and send it to anyone. If it can be decrypted using the Alice’s public key, one has a guarantee that the original message really came from Alice. – This use case is also called “Digital signing”.

The secret key is Alice’s crown jewel that she must not lose or share with anyone, because anyone in possession of the secret key can do what Alice can do. Well that is why it’s called the secret key.

Again, all of this can stay in the background and as it can be done for you by software on your computer/phone/…, that has access to your locally stored secret key. If you are interested in implementing some of this stuff in your life, you might want to take a look at

Just by reading this article and by understanding it, you should have a pretty good idea about what level of protection encryption can give you and where potential problems may lie. Again, never forget: your secret key is your crown jewel that you must never share and that your software must never leak. By the way, this is something that is especially hard to check with closed source software and also something that the heartbleed bug may have caused.

HR – follow up

I’ve started writing about it so I should finish it.

Where we left off: I had a job interview last week and I was expecting an offer in the next couple of weeks. Guess what: I received an offer today. And it did not come via a professional employer organization. No: Infineon made a valid, legit offer that truly is worth a thought.

In terms of money the offer is neither particularly good nor exceptionally bad.

Knowing what kind of salary to expect is one of the hardest things. Nobody in Austria really talks about their salary so you mostly rely on what friends tell you, internet research and what public research institutions pay (which has to be made public, because of transparency). Here are some examples of how I evaluated Infineon’s offer.

I have friends (also PhDs) working at Siemens VAI who earn about 5% more without the all-in option that Infineon has in all its contracts. The all-in option in Austrian contracts means that overtime will not be paid, as the salary is said to cover any overtime already. You would have to fight really hard to get rid of that option and accept a reduction in salary of about 15%. As 5% is not that much, it really comes down to the question of how much overtime you are expected to do. -> Neither good nor bad.

The post-doc position that my friend Reinhard has at university comes with exactly the same salary (according to the collective contract of the Austrian university system). My adviser also offered me a position as post doc, but I don’t think that the university landscape is for me. At any rate: -> Neither good nor bad.

Some information about what PhD graduates earned in Germany in 2012 can be found at http://www.ingenieurkarriere.de/_library/content/download/obj2407_Einstiegsgehaelter_fuer_Ingenieure_2012.pdf – Being  just a little below what was the 50% mean tells me: -> Neither good nor bad.

It comes down to this: I will not sign that contract right away (as I would have if the offer were really outstanding) but I am also not going to turn it down. Starting in mid-March I will start looking for other options and when I have a range of offers, I will sit down to do my pros and cons list. The fact that I could put to use my knowledge in the probably most efficient way will be on the pros side and the salary will not be everything as long as it is above my personal limit of … well I am not going to share that number.

Actually I will not share any other information about my job search on this blog from now on. With HR being good at googleing these days, I guess it is safer that way. I am still glad I blogged about this first job application as it helped me bring order to my thoughts, realize what is important to me and define what my limits are.

My Job #2

In my last post I claimed that part of my research is paid for by Infineon, the rest of the money is actually coming from public funding. That is why we have to provide an annual report about our activities once a year. Today I spent most of my time preparing the 2013 report.

The afternoon mostly went into updating my CV and preparing job application material. This week’s discussions made me understand that I should at least be able to hand in something at any given time. – You gotta be prepared ;)

bewerbung