Our Robotics adoption matters series heads away from the factory floor and moves outside for this instalment. Researchers Marc Carmichael and Sara Wilkinson discuss their Wallbot project.
You may have noticed more and more green infrastructure in our cities. As well as making the environment more appealing to residents and visitors, research has shown that green infrastructure, such as vertical gardens, green walls and facades, can have positive environmental and health effects.
When implemented carefully, GI can reduce energy and resource costs dramatically, and act as a carbon sink to mitigate the risks of climate change.
Cost is a significant barrier to the adoption of green infrastructure, in particular the recurring cost of regular maintenance and upkeep. This is exacerbated in major cities, where space is a rarity, and architects and developers are looking vertically for opportunities to introduce green infrastructure.
Utilising sides of walls and buildings for green walls and facades is becoming a popular approach to greenifying cities, however, this obviously presents a challenge when it comes to regular maintenance.
Workers are required to work at heights, and even to abseil down the sides of high-rise buildings so that plants can be maintained. Reducing the cost and OH&S risks associated with this maintenance, even just partially, may significantly increase the development of green infrastructure.
The aim of facilitating increased green walls led us to investigate how technology could be applied to reduce barriers to adoption.
A proven solution for the inspection and maintenance of infrastructure that is challenging or costly for humans to carry out is robotics.
Robots allow sensors to be walked, crawled or flown to locations difficult or dangerous to reach by humans, allowing measurements to be collected remotely and fed back to workers for assessment.
Sophisticated systems may even be able to perform maintenance operations autonomously, completely removing the need for humans to be put in high-risk situations. Applying robotic technology to green infrastructure would allow regular inspection and maintenance to be performed while reducing human requirements.
Developing a robotic system for the maintenance of green walls seems like a no-brainer.
The Green Wallbot concept:
To test the concept of a robot that can maintain vertical gardens, our team at the University of Technology Sydney (UTS), funded by a City of Sydney Environmental Performance Grant, has developed the Wallbot.
Early in the conceptual design stage, and based on discussions with manufacturers, maintenance teams, developers and other stakeholders, it was realised that green infrastructure comes in many sizes and forms. Solutions for automating maintenance would need to be tailored according to the requirements of the installation.
On the UTS campus we have several green walls of various sizes, ranging from metres to tens of metres in height. Then, just across the road, we have One Central Park, a 34-storey building covered in living plants (plate 1). Practical solutions for automating the monitoring, inspection and maintenance of plants needs to be suitable for the installation.
As a starting point, we decided to develop a solution using actuated ropes to manoeuvre the Wallbot across the green wall. The versatility and physical simplicity made this method of movement appropriate at this early stage of development.
For smaller installations, not requiring gantry cranes, BMUs, or other significant infrastructure to be installed on site is considered advantageous. Furthermore, once developed, the technology could be adapted to make use of such infrastructure if better suited to the GI installation, or if already available.
The trade-off with a rope-actuated robot was one of the many technical challenges this design introduces. Simple movements of the Wallbot require careful coordination of winches that extend or retract the ropes accurately on demand (see plate 2).
To facilitate this, we utilised commodity automotive winches, which were modified to enable closed loop control of the rope lengths and speeds. These “smart winches” are connected to a computer, and using mathematical models of the system the desired motions of the Wallbot can be achieved.
Wallbot includes a suite of sensors for measuring plant health and constructing a 3D map of the wall and surrounding environment.
A pair of 3D-sensing cameras provide both mapping and pose estimation capabilities, allowing the robot to sense its environment in 3D and estimate its location on the wall. This is particularly useful, given issues with the ropes stretching or winding unevenly, allowing these errors to be corrected.
An additional Normalised Difference Vegetation Index (NDVI) camera provides a non-contact measure of plant density from which we can infer the health of the wall. NDVI cameras are commonly used on farms where drones take aerial surveys of crops.
Given the proximity of the Wallbot to the plant during operation an NDVI camera with wide-angle lens was required. Combined, these sensors allow the Wallbot to create a 3D representation of the plants which can be used to monitor and log their growth and health over weeks, months and potentially years.
We are still in the early stages of the project and have many more developments to make. Due to COVID-19 restrictions we have been unable to perform site trials, however, we are eager to let the system out of the lab and on to the sides of buildings.
The next stage of the project is to refine the hardware to create Wallbot II, allowing it to operate in a variety of weather conditions, and to fully evaluate the benefits of the system in maintaining green walls. Longer term, we envision a future where robots such as the Wallbot are commonly used, helping us to maintain the plethora of green infrastructure that will be thriving in our cities.
We are currently looking for partners interested to help us in the next stage of the project.
For more information about the UTS WallBot project, please check out this short 2 minute video of the Wallbot in action below.
Marc Carmichael is a Senior Lecturer at the University of Technology Sydney (UTS), School of Mechanical and Mechatronic Engineering, Centre for Autonomous Systems.
Sara Wilkinson is a Professor at the University of Technology Sydney (UTS), School of Built Environment, Centre for Informatics Research and Innovation.
@AuManufacturing’s Robotics adoption matters series is brought to you with the support of the Advanced Robotics for Manufacturing Hub.
Subscribe to our free @AuManufacturing newsletter here.