SymMotion: A Novel Human-Aerial-Ground Vehicle Collaboration System With A Seamlessly Integrated User Interface

Team members

Ma Yuchen (EPD), Chung Wah Kit (EPD), Wong Wen Kang (EPD), Yeo Wee Hian Sean (EPD), Huang He (ISTD), James Raphael Tiovalen (ISTD)

Instructors:

Kwan Wei Lek, Norman Lee Tiong Seng

Writing Instructors:

Delfinn Tan

Teaching Assistant:

Congjian Lin

Currently, the cleaning of high-rise structures, such as the Supertrees at Gardens by the Bay, is costly, labour-intensive, and poses risks to workers. There are drone-based solutions available but they require highly skilled drone pilots, which is not ideal due to the potential manpower crunch brought about by the COVID-19 pandemic, ageing population, as well as the high cost associated with salary and training of personnel. Hence, in this project, we designed and prototyped a tethered drone system with a unified easy-to-use user interface. This enables semi-autonomous aerial-ground collaboration under one-man control to reduce the cost of manpower, cost of inefficiency, and safety risks.

Problem Area

Usage of micro aerial vehicles for cleaning and inspection of high-rise structures with non-planar facades in a built environment.

Current Issues

High risk of height-related accidents and fatalities.

High manpower requirement with operation certification.

High cost of maintenance and training.

Solution

The following UAV and UGV would be connected by a tether that would carry power, ethernet, and even water.

Unmanned Aerial Vehicle (Hawk)

You can inspect the 3D model of our UAV here!

Unmanned Ground Vehicle (Husky)

You can inspect the 3D model of our UGV here!

Simulation

A simulation was created to test and develop the various packages that would be used in the collaboration of the UAV Hawk and the UGV Husky. In addition to that, the simulation could also be used to train operators in using the system.

User Interface

The intuitive user interface allows single-person operations with operators performing one-click drone control. This removes the need for a highly-skilled drone pilot!

You can watch a demo of our user interface here!

Features

Relative Positioning

ArUco markers and ultra-wideband beacons are used to determine the relative positioning of the UAV and the UGV.

Dynamic Navigation

Dynamic obstacle avoidance and real-time path planning allow both the UAV and the UGV to avoid obstacles while performing their tasks.

Simulation

A simulation environment for our use case of cleaning Supertrees were developed to assist with the development and testing of both the UAV and the UGV.

Impacts

Safer workplace operations.

Reduce required man-hours by at least 85%.

Decrease costs by at least 65%.

Quieter operations.

Future Work

Better tether implementation.

The tether should be retractable so as to not obstruct the sensors and restrict the movement of both the UAV and the UGV too much. This would allow the system to achieve better relative positioning and localization accuracy.

More integrated collaboration algorithms.

More advanced algorithms such as Simultaneous Localisation and Mapping (SLAM) on the UAV or collaborative environment mapping would allow for greater autonomy and flight control assistance for users of the system.

Stronger processor units.

Installing stronger processor units on both the UGV and the UAV would allow the system to perform more expensive computations and process large video streaming data without sacrificing performance.