Smart Waste Management built and tested in VR

Share on facebook
Share on twitter
Share on linkedin
Building a smart solution to trash management issues was a challenge our interns proudly delivered a working prototype for.

What is the future of trash?

The future of trash depends on the autonomous smart waste management solution.

Rather than defining usage patterns in a deterministic manner, we believe the trashcans themselves need to become aware of their environment and the usage patterns surrounding them. In our evolutionary mindset, we see waste management completely automated using Internet of Things (IoT) devices. This is why we had the following internship assignments:

Why is smart waste management important?

Smart Waste Management is important to ensure our resources are properly allocated. Smart waste management makes our quality of life as it should be without too much adaptation required.

Overly full trash can causes streets to be littered with garbage.
An overly full Trash Can. Soon gone thanks to smart waste management. (Copyright: davidlida.com)

By making it easier to recycle our waste, we tackle the issue of not sorting our waste properly. For example: someone wants to throw away an empty plastic bottle, but the waste bin is full. So the easiest route for this person is to throw it in the nearest residual waste bin. The problem for this person is sorted, but henceforth, the sortability of this particular trashcan has become problematic.

Finding a fast and efficient way to recycle our toxic trash is something everyone benefits from. By automating the waste collection process, we can make sure that the sorting process will be respected right from the get-go.

Can IOT transform our cities?

Internet of Things can transform our cities, especially when it comes to effective waste management. Effective waste management starts with proper sorting habits of the city inhabitants. For that reason, making it as easy as possible for people to correctly sort their waste is one of the motivating factors to invent a smart waste management solution.

The reasoning for automating the waste management techniques is: if there is no obstruction to sort their waste (such as a overly filled trashcan), then it should become the easiest route to take.

Image recognition deployed to detect and count waste.

What are some revolutionary ideas in making a city ‘smart’?

Detecting and counting the waste can be stage one. Registering if a certain threshold is reached can be stage two. Next it should evolve to a stage where an autonomous self-driving vehicle is deployed to properly collect the waste. Additionally, making sure that the waste collection happens in the utmost smart way possible is the cherry on top.


What projects can I do in IoT?

Clean your room first

Jordan Peterson famously said “clean your room“. By this he meant that if you want to change the world, first make sure your own room is in order. Only then can you properly allocate resources to the things you deem improvable outside of your immediate sphere of influence.

Thus, in our project, “clean your room” means that we should first test our smart waste management solution in our own offices. Only then should we build a nationwide smart waste management solution.

In other words: we need to build a working proof of concept first. A prototype, if you will. This requires three stages:

  1. Building a smart trashcan.
  2. Building a self-driving vehicle.
  3. Enable communication between them.
  4. Test our contraption (first in VR, then in our offices).

1. Building a smart trash bin

A smart waste bin in our case knows when it is full (or rather: past a certain threshold) thanks to sensor data. It can deliver a data-driven impulse that it needs to be emptied. In addition it should also be able to predict when it will become full in the near future.So that it can be emptied preemptively when the waste collection robot is en route.

Is there a sensor to detect trash?

Detecting the contents of a trashcan can be built quite easily using an Arduino and a set of ultrasonic sensors.

Trigger and Echo of an ultrasonic senor for building the smart trashcan: https://iotify.help/virtual-lab/smart-trash/
Source: iotify.help

The ultrasonic sensor mounted on the lid will face downwards and send trigger waves towards the bottom of the trash can. In case the can is filled up to a certain level, the waves will bounce back as echo which will be captured back by the sensor. Sensing the time difference between the sending of trigger signal and reception of echo signal provides a reasonable measure of the level of the trash can.

Source: iotify.help

Our custom application uses data generated by three sensors. It has to use an uneven number of sensors to determine when half of something has passed.

Building a smart waste management solution using IoT technologies at ToThePoint.
Reading the sensor data with an Arduino and sending this data through a MQTT to a Java-built backend.

To measure it more correctly (and to prevent the deployment of the self driving vehicle needlessly) we use the mean number of the measurements of the latest ten measuring points. If five of those measurements are above our threshold, the bin is “full” and an impulse is sent to the autonomous vehicle. In so doing we can more accurately determine when the trash bin is “full” and needs to be emptied.


2. Building a self driving vehicle

A self driving vehicle in our case has to process incoming sensor data in real-time and send it over to the driver input management system in a timely fashion. This need for data means our self driving vehicle needs sensors, and a lot of them.

A sensor to detect obstacles (and a way to avoid them) has been created and tested in real life by our Intern, Joris.

Using the Bluetooth beacons to have our self-driving vehicle determine its own location has proven to be quite a challenge. Joris is still working on that. In the meantime we’re focussing on the RFID tags and a working MQTT.

We’re still a bit away from a fully autonomous robot as shown in the Tesla Gigafactory, but we’re getting there.


3. Enable communication between the smart waste management and our self driving vehicle

The next step is now to build a backend component which accepts, debounces and transforms the sensor data, and passes it along to a Kafka broker in the cloud.
Another backend component, a Kafka consumer, is currently the most prime candidate to host the trashcan-emptying logic, deciding upon the need to trigger a ‘time to empty me’ – alert (as a Kafka event). In fact, those events will be the actual triggers for the TurningPoint vehicle (https://tothepoint.group/news/turningpoint/) – on which Joris is working – to kick into action and take the trashcan to a collector station to be emptied.


4. Test the smart waste management in Virtual Reality

Robin and Lino have re-created the ToThePoint office space in Virtual Reality so they can train an artificial brain without distracting the rest of us. This smart agent will serve as a brain for the self driving vehicle.

Previous updates from Robin and Lino:


Smart Waste Management internships are in their final days

Smart Trash Management Interns Bert, Joris and Robin showcasing the working prototype of their autonomous self driving smart trashcan.
Bert, Joris and Robin in their trusted environment: the ToThePoint Labo.

Robin, Lino, Bert and Joris have only a few weeks left at our place before they go back to school.
We asked them what their experience was during their time here and we’ve got some quotes for you:

I will remember the homely atmosphere at ToThePoint. Right from the get-go, I was amazed by the informal atmosphere that Steven, the Upper Pointer, manages his team in. “Sir is not home”, is one of his famous taglines that I will never forget.

Bert

For sure there were directives, but it felt like we had all the freedom in the world to get it done in the best possible way we could think of. We were encouraged to challenge the proposed directives and suggest a better method if we could motivate it thoroughly. This way we felt the ownership and responsibility we needed to finish our assignment the best possible way.

Robin

I was amazed by the help I received from my mentor, Steven. Wanting to finish something at the end of the day on friday night. I didn’t know who to turn to but Steven, our Upper Pointer. He stayed with me until we found a viable solution to solve my issue. We were both really happy and relieved when it finally ‘clicked’ and our solution worked. The feeling of being able to simply walk up to your “boss” and get the help and support you need is priceless, if you ask me.

Joris


Also published on Medium.

ToThePoint

Leave a Reply