10 – My Secondment in Microlise: a new experience

Hey crew,

Are you alright?

Are you ready for reading a new post? Alright, let’s start then!

Almost a month ago I started my secondment in Microlise Ltd. As I already mentioned in the past: Microlise is one of the two companies that is involved within my project. They are experts in truck fleet management and they collect and analyse data from large truck fleets for helping their customer in maximising utilisation, increasing efficiency, and improving economy and safety for their fleets.

Fig.1 – The Microlise Ltd logo.

How is it going here in Microlise? Well, I have to say that working in the company has pros and cons. The cons are mainly related to the timing: I have to wake up at 6:00am every morning and it takes 1 hour-ish by bus to get to  Eastwood, where Microlise is located.

Then, I go back home very late (sometimes I come back home after 6:30pm).

Fig. 2 – No crew, it is not that terrible! ;)

This means that quite often I’m tired, and that when I have French and German classes at Uni I must rush in order to be on time after having a quick shower and eating something on the fly at home. It is a bit stressful, but that’s also life.

However, there are also many pros in working here. For example, interacting every day with my new colleagues I’m learning new stuff about Statistics and coding. Here, many people are real experts in these topics (as it is basically their job), and thus the project has suddenly speeded up. In fact, I learnt that different perspectives often helps in quickly and easy solve issues that seemed to have no solutions before.

Last week for example, I received new data containing measurements about the vehicle gross weight. It is a huge improvement since, with this new data, the accuracy of the fuel consumption model sensibly increase. I mean, we are still far from being able to consider the project concluded, but recent findings give certainly more confidence in the used approach.

Artificial Neural Networks (ANNs, do you remember? I talked about them to you in my previous post here), Random Forests (RFs) and Support Vector Machines (SVMs), three different ‘Machine Learning’ regression techniques have been recently implemented in the code. Thus, new interesting results are expected to be found out soon (I have a meeting set with my Supervisors for next Friday, I will let you know how it goes afterwards).

You never heard about Random Forests and Support Vector Machines, right? At the moment just keep in mind that they are Machine Learning statistics as ANNs and that they are cool (Very cool crew! Trust me! I’m an Engineer J). I really hope that I will be able to release a new post soon and I will give you some more information about these two methods. In fact, they are not completely clear to me either at the moment.  ;)

Other, pros? This week I talked with Mohammad (my Supervisor in Microlise) and apparently there is a possibility to collect some more data soon (more accurate and collected with higher frequency) for the validation of the models that we are developing. That’s great news crew!

More? No, that’s all, I’m sorry. Haha

Fig. 3 – A picture that I took this morning in Microlise.

No news also regarding the articles that I’ve recently submitted. However, in the meantime a conference paper that I wrote together with my friend (and ex-Supervisor in EMPA in Zurich) has been released. It is about “” and it was presented at the Accelerated Pavement Testing International Conference in Costa Rica last 20^th September 2016 (you can have a look at the proceedings of the conference here. The mentioned article is at pp 509-521).

So crew, at the moment that’s all!

I really hope that you all enjoyed this post and I will try to write a new one very soon to keep you up to date about the progress of the project.

As usual: just stay tuned!

Cheers,

FP13