After we’d passed the first hurdle on our path to MADhack hall of fame, we had quickly established a few communication channels. As I refuse to use Facebook for any longer than I absolutely have to, I went on and convinced my teammates to switch to a better and slightly more libre alternative – Telegram. There we have been in contact on an almost daily basis and we’ve shared a lot of interesting ideas, fun thoughts and some downright crazy notions. We’ve split into two groups, Max and Andrej went on to work on the cloud integration and mobile app; Adam and I took care of the hardware. Here I must commend Adam, because without his tinkerer’s expertise and a heart of a proper hardware botcher, we would have had to leave the competition. He’s designed the prototype and with his trusty 3D printer with a 1mm nozzle, he’s printed all the mounting and cable management components. As he was the youngest (at the start of the competition a ripe old age of 17), we’ve entrusted him with Resco’s money and all the important design decisions. He’s ordered everything that we thought we might need, while still fitting inside a very reasonable budget.
After the first week, during which we’ve collectively picked components and materials to order, we were forced to wait on the supply chain to serve us. When most of the stuff finally arrived, we could start working on our first prototype. The plan was, to have a smart weight sensors attached to the garbage collector truck, with RFID tags stuck to the garbage cans to pair the waste to the house it belongs to, save the weight value and some metadata to EEPROM and once it arrives to it’s station, upload the batch to Resco Cloud.
As a proof of concept, we’ve created a personal weight with GPS and connection to Resco Cloud over Adam’s Wifi Hotspot.
Wifi connectivity, GPS positioning and RFID reader were part of separate module.
One nice aspect of using GPS positioning is that you can use it as a source of real time clock. Unfortunately, when we polled the GPS for current time, it replied with an absolutely accurate time, albeit with date 30 years in the past. This slight oddity was easily rectified by adding 9.461e+8 seconds to the timestamp. To debug and play around with the GPS module, we’ve used this beautifully cinematic-looking software.
When we had the prototype somewhat ready, we’ve planned a trip to the village of Podkonice. Here we’ve met a bunch of lovely people, some of whom have called the mayor to query, why there were strange young guys measuring waste bins in their village. We’ve also noticed a slight problem. Despite being told by the former mayor that the villagers have standardized waste bins, we’ve discovered, that other than the typical tin waste basket that we’ve ordered, we’ve seen at least 5 different types and sizes of waste bins, each with a slightly different point of touching the lifting forks.
There was also another issue. We’ve thought about placing the weight sensors, where underneath the rubber thingies, but when the fork lifts bigger types of waste bins, the rubber folds underneath and scrapes against the ground.
This had the unfortunate consequence that our plan to add a smart weight scale to the lifting mechanism was no longer feasible.As we were thinking of a solution to this issue (one of the brainstormed ideas was to hook into the control unit of the truck and trying to calculate the weight from the hydraulic pressure), we were faced with another unfavorable fact. The lady clerk, who’s responsible for waste management of the village was less than thrilled by our presence, thinking our plan would only result in more bureaucracy and very little motivation for the villagers to sort waste and recycle. Disappointed and demotivated, we left the village. On the way home, we’ve decided not to give up and came up with alternative plan. Instead of measuring municipal waste, we would focus on sorting waste and instead of locals saving a few cents per month on lower taxes, we would use the principles of gamification to motivate them to recycle. Instead of modifying a truck, or hundreds of small waste bins, we would target the couple recycling areas in the village, and add smart weight scales to them. Adam went on to modify the schema accordingly.
To present our new idea, Adam has whipped up a simple visualization of what it would look like
The idea is simple. Villagers that want to participate in the game will get an RFID key-chain, Then they will collect points, which would represent grams of waste they sort. The players with most points would be declared the winners after every month, and optionally get some fun perks, like free beer in a local tavern, etc.
After consulting with Resco employees, we’ve started the work on the new prototype. We only had a couple weeks left and none of us had much time to spare. Andrej and Max are graduating from high school, Adam has a part-time job and I sleep on my daily commute to keep my sleep deprivation in check. Luckily, we could reuse some of the work done on the first one. We’ve decided to use whatever material was available to save on time and budget. First, we’ve made the weight scale ramps out of wood.
First part of the ramp didn’t survive Adam’s rigidity testing (drop kicking unfinished prototypes seems to be a favorite pass-time), but it shown to us, that we had to rethink the mounting of the pressure plates. After a few hours spread out to several days, we had the weight scale ramp ready.
While finishing it up, we’ve discovered a problem in our plan. Unfortunately, the ESP8266 doesn’t seem to have an i2c slave library available for it, thus we had to replace one of them with an Arduino Nano a few hours before the deadline. The whole prototype is a big ball of messy cable management, bad design, huge tolerances, copious amount of hot glue and sloppy soldiering. Nevertheless, it was a rewarding and fun experience. Tune in on the next post to learn how it all turned out.