We are Beach Cleaning Automated, a team of Mechatronics Engineering students from the University of Waterloo working on our capstone project. Our project involves the development of an autonomous sand beach cleaning solution that can eliminate waste from beach fronts before it enters bodies of water. Upon contact with water, this waste causes life threatening long term effects for humans, marine life and the biosphere.


Current solutions to the beach cleaning problem are manual, oversized, inefficient and costly. Our vision is an automated solution to beach cleaning that minimizes the level of human interaction required.


The ABCD is composed of four subsytems, each uniquely designed and retrofit for the application of beach cleaning.



Sensing & Controls

To enable autonomous functionality, certain sensing and controls features had to be developed. Firstly, the task of identifying garbage in the environment and differentiating it from obstacles had to be performed using a vision system. For processing this visual data on a mobile unit, hardware capable of quickly processing machine learning models had to be utilized.

For capturing vision information of the environment, a Microsoft Kinect sensor was selected. This was used because in addition to visual data, it also provides a depth map which can be used to localize features in the environment. For performing computations, an NVIDIA Jetson TX2 unit was selected as ideal. This is because of the inclusion of an onboard GPU with support for NVIDIA CUDA (Compute Unified Device Architecture)


For traversing the sandy terrain of a beach, a powertrain with sufficient torque, traction and speed was required. The minimum speed had to be slow enough that it could safely move through the beach and collect garbage. In addition to this, it had to have a maximum weight rating that supported at least 20kg. This was the approximated weight of all the components that would be mounted on top of it. Finally, it had to fit in a specified footprint of 1.8m x 1.2m.

To meet these specifications, a 1/5 - scale RC monster truck was selected. This monster truck was converted from using a gas combustion engine to use a motor and a battery. The powertrain was capable of carrying the required weight of 20 kg and was also within the footprint limitation.

Waste Intake System

As the ABCD drives along a sand beach, the top centimeter of sand is piled onto the robot's subsurface plow. The subsurface plow is waterjet cut from 6061 T6 Aluminum and acts as a shovel. A motor powered bristle roller then sweeps both the sand collected on the subsurface plow, and any garbage collected with it, up into the Waste Transport System. The roller is made of PVC tubing with four rows of Nylon PA6 Bristles used to sweep the sand.

Waste Transport System

The transport system is made of a custom designed wire mesh conveyor belt with slats. The wire mesh of the conveyor belt allows the sand to be sifted away and fall back onto the beach, while the slats of the conveyor belt transport garbage up the conveyor. The conveyor belt is made of stainless steel in order to ensure that it can not be damaged by the collection of sharp objects. The mesh is a tessellate of triangles that allows the collection of grabage as small as 1 inch x 0.5 inch. At the end of the conveyor belt the garbage falls into a waste storage bin where it can be emptied when full by the ABCD owner.

Meet our Team

We are a team of fourth year Mechatronics Engineering students from the University of Waterloo.

Taran Ravindran

Mechanical Lead

Johvonna Murray-Bradshaw

Project Manager and Systems Integration Lead

Misha Krutiy

Powertrain Lead

Bankole Adebajo

Sensing and Controls Lead

Collectively, our team has professional engineering experience in over 15 industry leading companies. The integral roles we’ve played in these organizations has made us well versed in applying academics within industry to create state-of-the-art technology.

Contact Info

Please feel free to reach out to us below or send us a message

Send Message

Your message was sent successfully.

Sorry!Something Went Wrong.