Hi there! It’s been more than two months since I started my PhD and officially joined the AMVA4NewPhysics network by signing an employment contract with the INFN. It was a quite intense period work-wise: getting up to speed with my project, working on analyses’ codes, writing blog posts, starting some PhD courses, getting to know my co-workers, presenting in a workshop and much more.

There was also a non-negligible amount of bureaucratic procedures and paperwork. In this post, I will tell my latest administrative adventure, which were some “safety-at-work” courses (i.e. sicurezza sul lavoro). Yes, I agree, that sounds boring! Nevertheless, I invite you to keep reading, because as you can infer from the title that will not be all I talk about today.


Safety at work is taken seriously in Italy, at least from a formal standpoint. There is a large set of rules and regulations for companies and workers, whose non-compliance might lead to high fines or even custodial sentences. These laws also apply to public organizations as the INFN and therefore I am obliged to comply to them as their employee.

I am not going to enter into the detail of those rules, you should already know them in case you work in Italy, because every worker is supposed to get a minimum of 8 hours of safety-at-work training. In my case, safety training was composed of a 4 hour online course on general safety and another 4 hours of classroom course with safety specifics for the type of work I carry out (videoterminalista, which is the preferred Italian word to refer to people that work in front of a computer).

I had quite good experiences with online courses in the past (e.g. Udacity and Stanford Coursera machine learning courses, which I highly recommend), but given the topic of this course I had low expectations, if truth is to be told. The best way to describe the format would be to include some example course slides, but I do not want to get issued by the company which organizes the courses.

The course was divided in several sections, each one composed of a long set of “interactive” slides with audio, shown as Flash Player applets. The slides were rich in content and the course material was appropriate overall, but for going to the next slide you had to wait until all the text was read at a very slow pace and then click to continue. This made the 4 hour experience very frustrating, because you could neither read the slides at your own pace nor just listen to them in the background. After one hour, I tried to speed-up the Flash Player clock without success and then I started dreaming about developing a bot to click for the next slide. Three hours and one exam later, the nightmare was over.

The classroom specific course, organized last week at TIFPA (Trento Institute of Fundamental Physics and Applications) for a small group of INFN workers was quite the opposite. While there was some overlap of the contents of both courses, the safety instructor used many examples and promoted discussion of safety policies among us, which is much better for learning than listening to a monotonous reading voice. Given the scientific and technical background of the participants, there were some odd but funny discussions, including whether the probability concept used for risk evaluation was Bayesian or not. It was also a good opportunity to meet other INFN researchers and learn a bit about what they did.

There were some people studying the biochemical effects of proton therapy, developing detector electronics for new experiments, simulating nuclear physics processes and a theoretical physicist working on gravitational waves (GW) research.

After the course was finished we went all together to a trattoria to have lunch, and the GW researcher, who is an ESR of the Grawiton ITN,  told us that he worked with the VIRGO collaboration and explained the basic design of the experiment and its purpose. Rumours of discovery were already all around the net, so I asked “the question”, but of course he told me he could not disclose anything, but also that we would not have to wait much. He was right indeed!

gw _discovery
Signals detected with the two giant interferometers from the LIGO experiment. The numerical relativity fitted to extract the parameters, their residuals and the corresponding spectrograms are shown below the detected signals. Figure from their main publication.

In the unlikely event that you have not heard, the detection of gravitational waves from the merger of two black holes has been announced by the LIGO and Virgo collaborations, which is pretty awesome! The detected signal is solid and the masses and relative velocities of the black holes before and  after the merge can be obtained from an analysis of the data.

The discovery has high scientific value in itself, but for me the real breakthrough is that a new and powerful way to study the universe has been demonstrated, which might provide many scientific insights in the years to come. Before leaving, I would like to point you to two interesting resources related to the discovery. The first is a short video from PhD Comics which describes in plain words what gravitational waves are and how they can be detected with interferometers, written and narrated by Daniel Whiteson, who also is a AMVA4NewPhysics network participant.

Lastly, in case you are feeling the need to crunch some data, the LIGO collaboration has done a fair and fast job in opening the data from the discovery and even providing some basic tutorial on signal processing. I hope at some point they will provide the full analysis chain, including the fit procedure and all the tools required to reproduce the shown results. I really think that open data policies are the way to go for scientific collaborations.

In case you are interested, I might write a post here on the current status and future plans regarding open data from LHC experiments. Feel free to comment for correcting me, asking for questions or starting a discussion on a related topic!