Yang Yiming has been complaining about how frequently we get food from Red Box Noodle Bar. I really enjoy it, but he kept insisting on the fact they do not serve authentic ramen noodles.

And so we decided to go to this new chinese place, along with Wang Ruiqiu and Xu Xinnuo, where they serve handmade ramen noodles and “oh boy, wasn’t it great!” Yiming gave it a 8 / 10.

Today they mocked me by asking “out of all the meals I have in a week, how many are chinese”, and I really can not tell!
Now that I think of it, concerning the last ~15 days, I can only remember ONE dinner when I ate with a fork and knife…

Am I slowly becoming Chinese?

Seriously, with so many Chinese people around me I have started to pick up some basics of the language already! But the proper pronunciation is very, very difficult to get right.

I’ve also had the opportunity to discuss lots of different subjects with them. What impresses me the most so far is the way the western media manipulate reality, and how often they hide what’s going on at the eastern part of the globe - or maybe it is just me who’s very naive and ignorant…
I mean, when I used to watch television (a looong time ago), all that I ever heard about China was: poverty, cheap and fragile electronics, and smog.

Just take a look at the video below about Chongqing. Notice the dimension of the city, and how clean and organized it is for a place that hosts a population of roughly six times more than New York! Doesn’t it look great?

Ok, enough about China for one post!

Updates on robotics

Let me think: what happened since the last post?

The Principal of the University of Edinburgh visited the lab today. We showcased Valkyrie walking and everything went smoothly.

Hmmm… that’s all I can remember…

As for my thesis, I am currently working on a benchmarking program for inverted and regular Reachability Maps.
In a nutshell, I want to extend the latest sampling-based methods for humanoid robots in order to support end-pose planning and locomotion on uneven terrains. And by uneven terrains I mean not only steps with different heights, but also with different inclinations.

The problem is that the generated datasets during the offline preprocessing are suboptimal and there is no way to classify them in terms of quality. And that’s why I’m exploring the benchmarking: to be able to tell apart one dataset from another, and to find the constraints required to generate datasets with better samples, which would make the most out of the capacity available in disk to store these configurations.

Aaaand… I’ll finish the post like that.