Intern optimises life and algorithms in Tokyo

Living in company accommodation where everyone goes home at the weekends has meant Arjun has had to "get creative" socially

After reading a brilliant article by Sarah Constantine and optimizing algorithms all day at Mitsubishi Electric, I've been thinking about the trade between indifference and optimization.

To optimize something is to choose the best action to maximise some benefit – like picking the shortest route to work, or leaving home a little earlier to beat the morning traffic. If you remain indifferent, you are equally happy using any action.

The duality between optimisation and indifference occurs because for every dimension you optimise, you lose a degree of freedom. For example, if you commit to optimising your morning commute, you lose the (self-imposed) freedom to take the longer scenic route.

I bring this trade-off up because it seems to me that every decision in Japan is fully committed towards the optimization school of thought. In the office, at 8:30 a.m. a bell rings and the lights turn on to signal the start of the workday. At noon, the lights turn off so that everyone leaves for lunch and at 12:45 p.m. the lights are back on with a bell to signal the resumption of work. At 3:00 p.m. a song plays and everyone stands up to do exercises behind their desk. It's always the same song by the way. And on the last Wednesday of every month, the entire company leaves at 6:00 p.m. to enjoy dinner with their respective teams.

These kinds of optimisations aren't limited to the office either. The trains run with an unreal efficiency, there are marks on the floor to tell you which side to walk, when eating at a restaurant you your order through a vending machine and then give your ticket directly to the cook.

Generally, I land on the more indifferent side of things, so to live in a culture so committed to efficiency is incredible. I don't always get it right. I usually walk on the wrong side of the path, I sometimes line up in the wrong queue for lunch, I never have exact change ready. I am trying though. That's the benefit of experiences like this, to be fully immersed in a completely different culture, philosophy, way of living.

Once his internship has concluded, Arjun plans to spend some time travelling around Japan

So what's work like?

It's been a real delight to get my hands dirty with some genuine data science at Mitsubishi Electric. The data is ugly. It's full of missing values, junk readings, messy timestamps – exactly as it should be. It's real data after all.

I've had to wrestle, wrangle and tangle with the data to get it ready to answer the business problem. First, I forecasted energy consumption in buildings to gain insights into air conditioner usage. More recently, I've been using GPS trajectory data to predict when a user will return home so that their smart appliances will know when to automatically turn on.

All of these are pilot studies for the company, so there has been a heavy focus on reviewing literature in the area. My team has been a pleasure to work setting me clear goals and giving me plenty of autonomy.

Arjun: "I've had to wrestle, wrangle and tangle with the data to get it ready to answer the business problem."

Socialising outside of work has been a little difficult. I have been put up in a dorm with other Mitsubishi Electric employees, so when the weekend rolls around it's empty.

It's forced me to get creative, so on NYE I decided to go to Tokyo for a Reddit meetup.

Suffice to say it was more than successful. Such an eclectic group of people from so many different cultures all getting together for a night on the town.

This time last year I would never have imagined I'd be spending NYE at the back of a smokey Shibuya nightclub at 3 a.m, raving maniacally about the intricacies of Tyler the Creator's Igor to a bunch of strangers I'd just met over the internet. Absurd.

All this and it's only been a month – Imagine what another two may bring.