How to keep research alive (Remote work in the time of covid19)

First, I would like to preface this piece by saying that most of our work is related to simulation and data science. This allows us to work mostly from computers, which makes us particularly suited for remote work. If your work is mostly experimental, some of these tips might not work, however it is worth thinking about the long term. Moreover, experimental work needs design of experiments, communication, literature review, theoretical review and writing. Most of these points can be done while in self isolation.

Really the remote has four components in no specific order:

Distractions (turn off all notifications, avoid multi tasking)
Family life
Virtual collaborations
Feeling of isolation

For distractions, it’s important to create a place where you can work. And create mental space between working and free time. I personally drag the dinner table to a different position in the room. That way there is a change in environment - signalling that it’s work time. For Family, it’s important to communicate to the other members at home when you are at work and when you are free. Direct and explicit communication is key. Furthermore for family I have abroad, I have a Whatsapp chat with my family where we share the day-to-day going’s on and meme’s. I also make sure to make time to call my family every week.

Most of this article is about virtual collaborations and feelings of social isolation points. By increasing the quality of our virtual collaborations we will diminish the feelings of isolation. Probably this period will last 2/3 months. Keeping people motivated and engaged is worth pursuing. I also, realize that peoples priorities change when family members or community members are threatened and need to be tended to. Nonetheless, we also have students from different countries that are dealing with tremendous hardship, and providing structure in the form of work for five days is an excellent way to stave off mental health issues.

Our home base (slack)

We have been working with slack since I started with the Energy In Cities group. Mostly we used Slack for the following;

sharing a good link,
please review my work (document),
sharing a good paper
sharing a bit of code I’ve been working on, and
asking if you have time to speak in person.

My current view on Slack that it can go into hyper drive. Maybe the conversations can be longer and we can structure our thoughts a bit more. I have seen some of the channels really use the (simple formatting) and document sharing options to expand the communication beyond these above examples.

Conversations and meetings (zoom and Skype)

This brings me to conversations. We have started hosting a 30 minute stand up meeting from Monday through Friday at 10AM on zoom. Zoom is a virtual conferencing service that is great for these things. The way it handles multiple users in a conference/meeting is very intuitive. You can send a link and everyone can login quickly and get started. I have found the connection to be very reliable and it handles large groups very well. For one-on-one meetings, Skype is still useful, however the quick here’s a link that zoom uses is very powerful and makes it easy to use. The free version of zoom gives unlimited time for one-to-one meetings, and up to 40 minutes for n numbers of members.

Two points to make here is that most of our communication is through body language and para-linguistics which get muted during a virtual call. We need to make sure take pause more often and ask for feedback. This way we prevent things turning into a monologue and allow all members to voice their opinion.

Cloud/Cluster computing

Running large simulations and working with data science requires hefty computers. Most students can only afford a small laptop with limited memory and computing power. This is where the cloud comes into play. The cloud in our case is separated into - web services - and cluster computing services. For those academics in Canada, they can get access to Compute Canada services. These are world class facilities that provide software support and amazing hardware.

The cluster can be used for jobs that run for more than 30 minutes and need anywhere between 1 CPU and 100’s CPU’s/GPU’s to run. The cloud services are great to host something like a Jupyter Hub. This is a programming, data handling and graphics plotting environment. With some basic programming skills you can replace most of the software scientists use such as excel and origin (for plotting). We use Jupyter hub for python developement, but beyond this language there is support for other languages such as Julia and R. Jupyter hub is easy to install on the any cloud service and get going. You can pay for cloud space and install Jupyter Hub, which will instantaneously get you a fully working computer in the cloud. Those that work on simulations and data science this might be particularly useful as they can scale quickly to more space, more RAM or CPU’s and GPU’s when their work gets more sophisticated.

We have hosted our own modified version of the Jupyter hub to help our researchers develop their python code. This has become a large project within our group called BESOS. This is a cloud platform that we have constructed for building scientists and modelling experts to use. We have build add-ons that will be specifically useful for Building scientists and modelling experts that wish to use machine learning and optimization techniques. We are running it for other academics in the building space to use Free of charge. You can sign up now at besos.uvic.ca.

Lastly, Co-ops task management (Trello/Gitlab issues board)

To continue the develop and maintain, the different online services we take on approximately 3 coop students per term. Since the start we have used trello or kan ban boards to make sure to list tasks for co-op students. These kanban boards are also available on gitlab boards. The kanban board is split up into a couple of columns:

New requests ( Feature requests are gathered here, PI determines if it should be a task)
Task list (Coops find tasks to do and drag them to In process)
In Process ( here we find the tasks the coops are working on)
Blocked ( If there is a issue which cannot be solved by the student and needs input)
Review (After the task is done, the person that requested can review the work here)
Done ( Finally the work ends here, this is a list of tasks completed)

For an example: https://gitlab.com/energyincities/besos/-/boards

Refocus on reading/writing

During these times, when our normal resources are not available, we can still read literature and write about our work. I suggest having a daily journal to take notes and think through your ideas. As we have (forced) time out from our work we might find that we actually have great idea’s. These must not be wasted, and need to be written down. Another thing we can work on are those pesky papers that are 80% done, let’s to push them over the edge. And really complete the work. :)

Closing

I hope we all fair well during this difficult time. Please feel free to give me comments on this work. As the time passes, we all need to adjust to the new normal and I think my thoughts will change as we progress. Currently, I personally think the lessons we learn from this time might actually improve our lives in the long run. All the travel we do as academics is a large strain on the environment and we might not actually get the best out of it. I hope we find good ways and develop good habits, methods and procedures that will help us be even more productive, communicate better and strengthen collaborations at a lower cost to the environment.