The article series AI for Low Carbon Cities is a collaborative endeavor between Urban AI and the Lee Kuan Yew Center for Innovative Cities. This series embarks on an exploration of pioneering initiatives within Asia, showcasing the potent synergy between urban Artificial Intelligence and sustainable practices in the face of climate change. Through in-depth case studies, this series delves into real-world applications where AI is harnessed to shape Low Carbon Cities, offering innovative solutions that optimize energy usage, enhance transportation systems, and mitigate environmental impact.
ClearBot for Port Cities
Plastic is one of the most persistent pollutants on Earth. At every step in its lifecycle, even long after it has been discarded, plastic creates greenhouse gas emissions.
In our oceans, which provide the largest natural carbon sink for greenhouse gasses, plastic directly chokes and smothers a host of marine animals and habitats and can take hundreds of years to break down. As it does, sunlight and heat cause the plastic to release powerful greenhouse gasses, leading to an alarming feedback loop. As our climate changes, the planet gets hotter, the plastic breaks down into more methane and ethylene, increasing the rate of climate change, and so perpetuating the cycle.
At least eight million tonnes of discarded plastic enters our oceans each year. (Parker, L., 2021) Today, at least 60% of plastic marine debris are found within urban centers (Lebreton L. & Andrady, A. 2019) and their waterways would funnel the plastic waste to the ocean, impacting wildlife and biodiversity. More than two-thirds of these urban centers are port cities (Pavia, R. & Z, T. (2021) which constitute interconnected sea transportation networks for global trades and cruise shipping.
In port cities like Hong Kong and Surabaya, the plastic problem worsens due to the lack of an inadequate recycling system. With high population density and land scarcity, Hong Kong lacks the available land to provide an effective recycling system as the city must prioritize other land developments. In the case of Surabaya, it is one of the largest cities in Indonesia with a major port which engages in a high volume of import-export activities. The amount of trash flow caused by port and human activities in the city is found to be 5 times the collection capacity of the current efforts.
To address the severe problem of the plastic crisis, the team at Open Ocean Engineering with backing from Hong Kong University set out to build a solar powered floating drone, Clearbot, integrated with advanced computer vision programs to detect and collect trash from water bodies. According to Clearbot, the robot is 15x cheaper, has 5x more reach, and removes 2x more trash daily, in comparison to any current solution.
Clearbot is currently carrying out one of its projects at the land reclamation site at Shek Kwu Chau, Hong Kong. The area is an outlying island where the Government is reclaiming a portion of the land. The site requires a daily removal requirement of 3 to 4 tonnes of trash per day and the state has used barges, cranes and nets to remove floating waste. The overall process was found to be more expensive, tedious and slow. With the help of the aquatic robots, 250 kilograms of trash was effectively removed by each robot at a quicker pace within every 2 hours.
By incorporating machine learning, the robots are trained to differentiate between 64 different categories of trash, which include but are not limited to plastic bottles, plastic bags, milk bottles and food containers via existing data from photographs of trash donated by non-profit organizations and schools. New images are photographed by the robots and the newly captured images are then transferred to the cloud platform to improve the image-detection process of the robot.
All the data that Clearbot collects can be displayed on an AI-analytics dashboard, showing what trash is being collected in real-time while describing water conditions through the sensors integrated into the robots; producing qualitative reports and graphs based on AI-analyzed data and sensory information.
The plastic trash collected will pass through the robot’s open bow and accumulate inside the robot for collection. As soon as the robot’s battery starts to run out or when its trash can is full, it returns to the central docking station.
Clearbot is continuously working on its AI technology to remain competitive and sustainable as a business. With its machine learning abilities at the current accuracy rate of 95%, Clearbot aims to sustainably remove as much plastic waste from water bodies by expanding its services to more port cities.
After the trash is collected, the trash is sent to local disposal partners for sorting and an average 80% of the collected plastic waste is used for recycling. There is remaining trash which is unrecyclable as the materials are filled with contamination. To ensure the waste is carefully treated, Clearbot has chosen to collaborate with recycling partners who possess the skills and sorting infrastructure to treat and segregate the waste to contribute to the recycling value chain.
With the help of loop business action in its business model by recycling the waste for reuse, ClearBot estimates that it can save 2.25 tonnes of CO2 entering the atmosphere per robot per day by accounting for the collection of plastic waste and plastic waste recycled.
In face of existing competition from dredging companies as well as individuals who own second-hand boats and hire human resources to collect all types of trash via traditional nets on the sea, the limitation of Clearbot’s robot is that it lacks the functionality to collect plant trash. However, the company wants to remain focused on plastic trash clearance to protect water bodies and the environment. To stay economically viable, there is general reliance on government funding and support for Clearbot to effectively roll-out its projects to combat the plastic crisis.
Analyzing the robotic solutions of Clearbot, the computer vision currently only targets collecting on surface trash and up to 1 meter deep. Yet, studies found that plastics break up to microplastic fragments which are passed unchanged along waterways deeper into the ocean (Information Sheet: Microplastics in drinking-water, Who Health Organisation (2021). To strengthen Clearbot’s business foothold to attract more collaborations with government agencies, the expansion of new services to include capabilities to further submerge and detect plastic trash under the water will help to minimize more plastics being released into the environment. However, the research and development aspect of building a submarine robot is still premature for Clearbot and this demands more funding to support robot development.
The former version of the robot had included solar panels to power the robots. The current design only consists of battery storage as they found that batteries can provide longer battery shelf life. To foster a greener environment and reduction of a carbon footprint, Clearbot can consider building a hybrid battery model by including both battery and solar panels on the robot for interchangeable use to safeguard the increase of carbon emission.
The company can also build a connected network with other local stakeholders to advance circular economic solutions for the plastic waste problem. For example, in Indonesia, there are external players such as ReBricks which turn hard-to-recycle plastic packaging ranging from instant coffee, snacks and single-use shampoo into paving bricks for construction companies and individuals to refurbish their yards and gardens. According to ReBricks (Two women entrepreneurs turn hard-to-recycle food packaging into paving bricks in Indonesia, Channel New Asia (2020) are typically sent straight to the landfills by recycling companies as it is difficult to separate the different kinds of plastics and aluminum foil. In this case, Clearbot can tap in on the eco-friendly and green movement organizations for collaboration to ensure unrecyclable materials are being put to beneficial use.
Moving forward, Clearbot can also provide the solution for major cities along the coastlines. As trash can travel throughout rivers and oceans, it would end up being accumulated on the coasts. Plastic littering and poor waste management practices by humans living near the coastlines can also lead to the problem of plastic pollution. This is where Clearbot can step in to provide a cheaper and more efficient method of waste collection to clear the rubbish-choked coasts.
By Evonne Li, Elissa Gowika Hartanto, Lee Wai Loon, Anastasia Ejov, Iacopo Testi
