About a year ago, I've reconnected with an old acquitance, due to her installing Telegram and popping up as a notification on my phone. Since then, we've caught up on our lives and I've learned, that with her degree in philosophy, she has found a nice and cozy job at a mid-sized IT company in Bratislava, the capital city of Slovakia
stan. One day in the not so distant past, she has mentioned that her company was organizing a hackathon for high school and university students with an interesting focus on helping Slovak towns. I, being a politically and socially active being, got interested and and invited by her to attend. So here I was, on my way to one of my least favourite city in Slovakia, filled to brink with my raging imposter syndrome, knowing full well that I was about to meet people far above my coding competency. I reckoned, that even if I get myself a bit humiliated, I would at least gain some more experience, get fed on her company's money and have an opportunity to share a bottle of wine with an old friend. Having a new topic for my growing blog was a fine bonus as well...
MADhack is an annual hackathon (or is it... more about it later) for students and recent graduates, which aims to, apart from providing a decent PR for it's host, improve the lives of Slovak people by implementing smart tech in the local town governments. Sounds good, right? The company behind all this is Resco spol, s r.o., founded in 1999. From what I gathered (without doing any previous research), they make most of their income by offering their cloud services and solutions to companies and societies in the Western Europe and US of A. What I found interesting from the discussions with my friend was, how she found their philantropic efforts to be honest and true. I, being a card-carrying skeptic, may find it hard to chew, but in my time there could not find any proof of that not being the case. I usually view corporations as a single psychopatic individual, but I guess they are innocent until proven guilty.
Before I went to the event itself, I've decided to take them up on the offer of free accomodation in one of the local hostels. One indication that this weekend would be a lot of fun was, that even the male receptionist in the hostel recognized my Ubuntu merch, that I carry to every such event.
The hostel wasn't anything fancy, and I shared a room with bunch of people attending the same event. The bed was comfy and there was a bar filled with fun-loving Spanish folk in the basement, so I couldn't complain. It was also just a few tram stops away from the venue of the "hackathon".
When I got there, I've found out that most of the attendees came already in teams of 3-5 people and that I had to be in a team to compete. Luckily, I got in with a team of three very clever high-school students, and we've hit it off immediately. Once the attendee registration was finished, there was a short presentation concerning the company, the project Lepšia Obec (City Smart Services) and the details about this year's event. There also was an announcement that the finalists would get a significant budget to realize their project (€20 000) and that the winning team would get €5 000 prize and a trip to 2019's Resco conference in Rome.
Soon after we've formed a team, we've had to pick a task we wanted to tackle from a list. These tasks and problems have been brought forth by mayors:
Short term parking solution. The problem presented was that there was a system of free short-term parking, that the drivers would not honor. Thus, there is a need to measure how long do people stay parked in one spot and preferably capture the licence plates of misdemeaning residents. We did not pick this problem to solve, as we've collectively agreed that it was a bit boring subject with obvious solutions.
- Smart Lighting - The town already has ~300 smart city light, needs ~100 more.
- Illegal lumberjacking - Monitoring illegal timber operations and catching the culprits.
The second and third problems were assigned by one mayor, which reminded me of a relevant XKCD:
This problem had both the appeal of not being a (complete) privacy-eliminating nightmare fuel and that the mayor's reasoning was both economically and ecologically understandable. He wanted us to think of a system to monitor the amount of communal waste each of the resident households produce and to encourage them to adhere to the Five R's (Refuse, Reduce, Reuse, Repurpose, Recycle). Since the village of Podkonice pays the waste disposal company per kilogram of communal waste, decreasing this number would also decrease the spendings of the village, as recycleable waste disposal is funded by the recycling fee included in most of the electronics and other products. The mayor was even inclined to set a cutoff limit, under which the household would pay nothing for the communal waste disposal service, even further encouraging good behaviour. As it is a small community with close neighbourhood relations, he wasn't worried about people cheating the system, making the whole task easier and solvable.
Counting real population of a town. Official number is 2057, suspected is about 800-1000 more. Town needs to get information about why and how many people did not register as residents. We passed on this issue mostly because of our moral compass not agreeing with tracking individuals down to the place of residence.
Scaring stray cats and fornicating teenagers away from wholesome areas... Again, boring and easily solved by a camera, speaker and police doing their duty.
It is probably obvious, which problem my team has chosen to tackle. While we were picking the task we'd like to complete, we had the opportunity to talk to the mayors themselves. This was useful, as we could ask about various details, preventing misunderstanding of the assignement and come up with better solutions.
Apart from mayors, there were also employees of RESCO, with various specializations (such as cloud, hardware, ...), that we could bounce our ideas off of. This proved very useful, especially as none of us had any experience with Resco Cloud, and we've received a short but informative demonstration of it's workings.
Of course, a hungry techie is a mean beast, so it was a blessing that there have been a buffet with loads of food and drinks. It has been continously refilled with items appropriate for breakfast, lunch and dinner depending on the time of day.
The proposed solution(s)
After the initial joking around and brainstorming crazy ideas (we've contemplated a contrived solution to all the listed problems by means of automated paintball turret), we've set up to solve the issue at hand. I've quickly discovered, that Adam has great understanding of hardware and first-hand knowledge gained by working with Atmel-based boards. Fortunately, Max and Andrej both were quick to show off their talents and cognitive abilities while we were brainstorming our solution. With the goal of fitting within the €20 000 budget, we've worked out two alternatives. If we can get hands on access on the waste collector truck, we can modify it to weigh the garbage cans and pair the data with the household using metal-passive RFID tags. This variant would cost a few hundred euros per truck + €1 per garbage can. The other possibility is to fit a custom-made smart scale in the bottom of each garbage can in the village and although it is technologically more interesting, it would increase the total cost ten-fold. We've also done some napkin-math to calculate the estimated power draw of both solutions and figured out that given proper optimalization of our code, a dollar-worth of tritium would power the garbage scale for many decades, even accounting for the inefficiency of capturing the energy of the decay using photovoltaics... Unfortunately, even though beta decay of tritium is pretty much harmless, it may be a bit tough getting all the necessary permits to propagate nuclear-powered garbage cans across the country... I'll report back on the progress of the project at later date.
Is is a hackathon without any hacking?
You may wonder when the actual coding comes in. Curiously, there was hardly any coding involved in the event. There was a reason for that. My friend, Lucia, mentioned that in the past, the company has noticed that many young students would, when presented with a problem, jump at the first solution they thought of and try to implement it right away. By not asking the contestants to produce a working prototype in a couple of day, they motivate them to think through the problem properly. That way the solutions would be to some degree optimized and realistic. My impression of that was that it worked out about 60% as planned. Some teams, instead of spending the time on calculations, brainstorming, hardware comparisons and searching for suppliers, would spend the night producing the most eye-catching presentation. Fortunately, as far as I could tell, the judges were techies and could not be fooled by pretty animations, especially when some of the teams, ours included, has scrutinized and picked apart their ideas. Seeing as we've made it to the finals, it seems our approach of spending hours thinking and just minutes preparing presentation was in the end validated.
All in all, I can honestly say that I've had great time at this "hackathon" and was glad to meet many talented young people. I also look forward to working with them to improve the efficiency of local governments and services. Now we've got 4 months to work on our projects and even if we don't win the prize, we can in good faith proclaim, that we've done our best and hopefully inflicted a bit of positive change in our country.