Backwards Breakthrough - 25/07/18

Today was another meeting with Andrew, and again it was full of ups and downs, and thats not the drone! It started off similarly to previous meetings, being a very slow process with the hours moving too quickly for me to believe I'm actually getting anything done. As I had made significant headway in putting the drone spreadsheets together, I'd noticed that each section had between 30-60 commands... and considering there are 6 sections in total that adds up to a lot of programming. I had sent some drafts of the spreadsheets to Andrew for him and his interns to start working with before this meeting and so today Andrew started to show the progress... which was... slow. It took a long

Andrew and I attempting to tame a disobedient drone

time for the drone to perform the command to take off and rotate, and sometimes even if it did take off, it would then veer off in expected directions. This is because the drone needs to stabilise itself as it rises, and depending on the floor, for instance the black dance floor we have in the studios and in the theatre, is almost too black as there is nothing for the drone to recognise its location and hover above it. The amount the drone veers off to also changes every time, so there is no way to predict where the drone will end up. So if I hope to set up a starting position on stage, and drone veers away, it is then out of sync with its starting position and the rest of the programming will remain out of sync. This is very unwanted as the chances of something going wrong are now significantly higher. 

Whilst I was watching this happen and observing Andrew struggle to get the drone to take off efficiently, I panicked slightly and considered thinking of a back up plan. All that was coming to mind was reverting back to the playful exploration I had with the drone when Georgi watched, and all the interesting reflections she made. I mentioned this to Andrew and showed him the video and he also found it quite interesting, especially from an academic research point of view. Ultimately it was decided that programming the drone was going to take up too much precious time, which initially felt like a complete u-turn in my research. 

Andrew and I thus had a long detailed discussion about the theories behind my research and reminding me of my research questions which I think had become a little lost in my quest for programming. Andrew reminded me that drones are not designed to be programmed by code, and what happens with code is that, if we send a command for the drone to move backwards for 5 seconds, it receives the backwards command and starts moving before it has received the command for how long for, therefore the drone travels further then intended. Andrew called this a problem with transmission time and if we had months to complete this project, we may be able to find a solution, however, we have weeks, not months. 

On the subject of interacting with the drone naturally, Andrew had this to say 

"The beauty of it from an academic point of view is that you are interacting with it native, it is responding directly to you, in all other scenarios it is not responding directly to you at all. You have created a routine that you want it to do. If the premise of your work is the interaction of human and technology within dance, you already have a lot of information about the difficulties. The theory part is really interesting academically and all these problems that are occurring are all part of it and how you adapt."

This reminds me of one of my key research questions: Are machines able to organically create its own choreography? And by allowing the drone to react and move with me, I'm allowing it to do just that. Andrew reminded me of another one of research questions: How can we set paradigms for working with technology in dance in the future? By reminding me of the importance of setting boundaries.

"First thing you have to do when discussing autonomy and having machines think for themselves and robotics, is who defines the boundaries, how far can they go, in this case, how far can the drone go in distance, creates a physical boundary. But in the case of a military robot, how far is that robot allowed to go, is it able to make the decision to shoot someone? Its the same with autonomous vehicles, the autonomous car is built on the fact that you set it boundaries and you try to teach it the same boundaries that we have as humans; the white lines on the roads, the traffic lights, giving it all these restrictions to try and teach it to fit within our boundaries. And that is exactly what you're trying to do, and that is what you will overcome with this research." 

I hadn't realised that all the problems I am currently having are actually making my research all the more valid, I'm addressing my research questions without forcing the answers to appear. And when further problems arise, I need to remind myself of this and the progress I have made already although it may seem small. This was always going to a difficult process as I'm exploring things that don't always get the chance to be explored. Our lengthy discussion ended with this statement from Andrew:

"I think you are a lot further along than you think you are, because you started to detract from your research questions, because you are trying to perfect a performance as thats natural for you. Everything in dance is all about getting it right, practice practice practice until its faultless as a fault gets picked up. And you need to move out of the space as what we are doing here is improvisation, in an environment in which you don't yet understand the boundaries of. Don't focus on perfection and its the imperfections that answer your research."

And so, despite the slight upset in the intended research process, I feel as though I've had to move a few steps backwards to practically leap forwards. The discussion left me not feeling disheartened at the change in method, but feeling re-energised after discovering the achievements I have already made. I really feel as though I'm learning more and more every day and I'm starting to see that successes and failures are still valid progress. In fact, Andrew did warn me in the very first meeting that I would suffer from a lot of back and forth, and he was right, this drone flying malarky is proving to not be such a smooth ride after all!