What it does
Compact and affordable, KEVIN is a prototype focused on operational efficiency and cost-effectiveness. Iterations aim to expand its capabilities to rival commercial deep-sea data collection solutions, while upholding its core design principles.
Your inspiration
Underwater data collection typically involves three key steps: identifying targets of interest, surveillance of target organisms / regions and collection of samples for subsequent laboratory analysis on land. There are limitations to technical divers’ dive durations, depths and areas, and deep-sea environments remain scarce of data even with surface vessels collecting data from open seas. In comparison, the ability of underwater vehicles to sustain extended missions, operate at significant depths and handle dangerous missions sparked our team’s efforts to develop a specialised vehicle for oceanographic data collection and sampling.
How it works
KEVIN is a low cost and compact Autonomous Underwater Vehicle (AUV) capable of diving down to 100 m, deep enough to explore shallow and Mesophotic reefs. It is equipped with visual and acoustic sensors, enabling it to perceive and map both clear and murky water conditions. It also has a robotic arm with a soft gripper made of a flexible TPU material, capable of collecting fragile samples without damaging them. Kevin is also capable of placing research probes virtually anywhere with its magnetic probe dropper and self-propelled probe launcher. The probe launcher is completely self-contained and uses wireless charging to recharge probes, making it easily reusable. Measuring just 60 cm across, this vehicle is capable of navigating tight spaces and conducting marine research in places most humans cannot reach.
Design process
KEVIN started with a smaller prototype bot named Lucy. This bot is capable of just basic maneuvering and visual sensing. This served as a testbed for the basic functions of Kevin. Lucy’s shape was inspired by the shape of Manta Rays. These Rays are known to be highly maneuverable despite their large size, and we decided to borrow some of its characteristics and apply them to Lucy. After numerous sketches, we finally created a CAD drawing of Lucy with Solidworks and printed her out and installed the sensors and thrusters to test out her functionalities. Once we understood the basic characteristics of Lucy, we adapted her design and added additional bays for larger sensors. After a few iterations, KEVIN’s shell was born. The other subsystems such as the hydrophones, droppers and launchers were tested separately before being integrated with KEVIN. KEVIN was then submerged in our university swimming pool to fine tune his control and test out software to make him autonomous. KEVIN is still undergoing minor upgrades to this day as we continue to perfect him.
How it is different
The concept of autonomous underwater vehicles (AUVs) is not new, intensified efforts in recent years aim to optimise them for effective deep-sea data collection. However, drawbacks of commercialised AUV include the substantial costs to customise vehicles for scientific sampling and reduced manoeuvrability due to the bulkiness of vehicles that may interfere with sampling from targets of interest. Affectionately named KEVIN, our AUV is designed with a strong focus on operational efficiency and cost-effectiveness, currently equipped with core components for underwater positioning, acoustic communications, high-resolution imaging and basic object retrieval (sufficient for underwater debris collection) with the whole system only weighing around 20 kg. This serves as a promising headstart as we integrate more specialised samplers and sensors, advancing us closer to our goal of comprehensive sampling and data collection in the depths of the ocean.
Future plans
On the hardware front, KEVIN’s current modular vehicle frame features sensor bays that allow for new instruments to be easily and securely attached. The next phase involves integrating specialised sampling equipment such as samplers that collect environmental DNA from various depths as the AUV moves through the water column, advancing our capabilities in marine biodiversity and habitat research. Concurrently, the software team will focus on enhancing KEVIN’s autonomous path planning capabilities, enabling the AUV to identify and prioritise ocean targets, adaptively exploring areas of interest to fulfil complex marine sampling missions.
Share this page on