Developing a robot for collecting shoreline debris requires a lot of testing the robot in the field. Here is what we are learning in the process!

Shoreline debris is a massive problem. Over 16 million pounds of trash were collected during the 2015 International Coastal Cleanup. Food wrappers, cigarette butts, and plastic bottle caps were among the top ten items collected.

I was shocked to notice this first hand on Toronto Island, where there was a trail of debris that extended along the entire beach.

robotmissions_to_isle_debris

Being a part of the Maker Movement, my mind was instantly racing with ideas on how we can fix this!

With Robot Missions, we’re developing a robot platform to help make shoreline cleanup more efficient and use robots to reach difficult to access or dangerous locations.

robotmissions_robotwaterbottle

The robot is a 3D printed rover that is tele-operated with a nunchuck. Its brains consist of a Teensy for the motors, and an Arduino for the sensors and communications. Its arm is interchangeable with a sensor probe or claw.

robotmissions_fieldtest_snow

We go on Field Tests to understand how the robot performs in the environment. One of the most exciting things is when we get to discover emergent features. Our first Field Test was at Cherry Beach, and second one at Sunnyside Beach in Toronto.

Here’s a video of the robot during the first Field Test:

Behind the scenes there is a bit more that goes in to the Field Tests:

1. Preparation

robotmissions_conceptops

The first step is to get a group of volunteers together to help with the Field Test. We broadcast the event on Facebook, and make a map to show the location.

We create a Concept of Operations Guidebook that gives background on the location that we chose, the roles, and tasks. There are worksheet areas to record down observations and answers to thought-prompting questions. There’s also an agenda to guide us in terms of a schedule.

A few of the items in our Field Test kit include:
– First Aid kit (important!)
– Measuring tape
– Rope
– Tarp
– Gloves
– Duct tape

On our first Field Test we experienced sun, rain, wind, and even snow! Not only is it important to prepare the robot for a variety of weather conditions, but also yourself as well.

2. Arrival & roles

Next, we arrive on site. But sometimes, it isn’t as simple as that.

robotmissions_walktobeach

On our first Field Test, the park didn’t permit us to enter with the robot. We had to decide what to do next, so we navigated to another nearby beach. It’s important to be flexible and to have a communications channel in place for contacting everyone.

When we do finally arrive, we look at the location to find an interesting area to test the robot in. Some details we consider are:
– Various types of terrain (grass, dry sand, wet sand, twigs)
– If there’s a challenging trench to try to navigate nearby
– The amount of debris in the area
– Proximity to water and the inclination to it
– If we’re curious about a particular land feature

robotmissions_soiltest

Then we unpack the robot, and determine the roles that everyone will have during the Field Test. These can be:
– Robot Operators – controlling the robot
– Task Assistants – helping to set up and run the tasks
– Environment Observers – recording the effects of the robot on nature and its surroundings (and vice versa)
– Mission Operator – keeping track of time, checking in with everyone
– Documentation – recording the Field Test with photos and videos

3. Robot tasks

We have set three tasks to accomplish in the Concept of Operations Guidebook. These detail the objective of the task and some observation prompts.

Importantly, we do not detail how we are going to accomplish the task. We need to take into consideration all the factors in the field to understand more about the robot capabilities.

robotmissions_measuring

For example, in Field Test #2, the task was “terrain transition” – to navigate from point A to B. The path was through tree roots and rocks with dry sand. We learned that if the robot encounters something that resists one of its wheels from spinning as fast as the other, then its direction will change – and this error increases over time. In the above photo, the robot travelled 5.7m, with numerous twists and turns!

robotmissions_tennisball

One task is always about debris retrieval. Currently the claw is designed for medium sized debris. We’re able to learn:
– Accuracy of the robot
– Best approaches to driving towards the debris
– Gripping ability
We’ve been able to grab plastic water bottles, coffee cups, discarded tennis balls.

robotmissions_fixingbot

Sometimes we encounter flaws and the robot breaks, but we are always ready to try for a temporary fix with duct tape!

4. Observing the environment

robotmissions_ladybug

It is critical to be attentive to what is happening around the robot. As we were driving the robot, one of the crew members noticed a ladybug. We immediately stopped and began to reverse. We definitely do not want to harm nature.

robotmissions_seagull

It was interesting to watch as a seagull inspected the robot. It walked in a circle twice around the robot, in quite close proximity. It then tried to “jump” near it, possibly to scare the robot. As it turned to look at us, I can only wonder what it was thinking!

5. Debrief

robotmissions_debrief

This is one of my favorite parts, as we go through everyone’s thoughts and observations about the Field Test. We learned from the first Field Test that it’s important to do the debriefs on site, rather than back at HQ.

One memorable thought from a debrief session was when my friend mentioned about the boundaries between human civilization and nature. If you look at urban cities, this boundary is incredibly strong. But on shorelines, it becomes less pronounced.

These thoughts are incredibly valuable, and shape how we approach the next Field Test and the improvements that are prioritized on the robot.

robotmissions_robotinbush

The next step for Robot Missions is a Mission Pilot. The goal is to collect 5kg of debris from Toronto Island at the end of August. It will be similar to a Field Test in execution, but focussed primarily on that goal.

The outcomes from the Mission Pilot will serve as a model for how this effort can be replicated in their local communities.

We have a crowdfunding campaign where you can contribute to support us fund the Mission Pilot – check it out on Kickstarter.

robotmissions_envisionedrobot

The robot platform will reach a solid proof of concept design through iterating on it during the Field Tests leading up to the Mission Pilot. It will be released open source, along with a set of interchangeable modules. A stretch goal will be to replicate a robot using the plastic debris that we collect.

robotmissions_fieldtest_crew

Imagine fleets of these robots deployed on shorelines! It will bring efficiency to tackling this problem through day to day cleanup. It will be easier for them to reach areas that are difficult to access. Some will have different tasks, such as watering and planting grass seeds, with the major goal of preserving our natural spaces.

robotmissions_fieldtest_selfie

The broader purpose of Robot Missions is to enable makers and environmentalists to collaborate on improving the environment — it’s more than the robots. Robot Missions aligns with goal #15 Life on Land, of the UN Sustainable Development Goals.

It’s a big problem, but I believe that with the power of the Maker Movement, we will be headed in the right direction to make a difference.

Join the journey with us on Facebook and Twitter.

PS: Here is a photo a friend was able to snap of the Mission Pilot site, while they were flying on the way to Maker Faire!

robotmissions_aerialview_torontoisle_dagohh