ERU - 2023

Permanent URI for this collectionhttp://192.248.9.226/handle/123/22458

Browse

Now showing 1 - 20 of 23

item: Conference-Extended-Abstract
ERU Symposium 2023 proceedings 5th - 6th December 2023
(Engineering Research Unit, 2023-12) Gamage, JR; Nandasiri, GK; Herath, RP
item: Conference-Extended-Abstract
Investigating the design process of traditional costume penetrating a niche market
(Engineering Research Unit, 2023) Bhagya, KPN; Karunaratne, PVM; Ranathunga, GM; Ranaweera, HRAT
The design process defines each market segment in the apparel sector. Traditional costumes of Sri Lanka gained a long historical and cultural demand among Sri Lankan consumers. The most demanding costumes are "Nilame" costumes (male costume) and "Ohoriya" (female costume). This research identifies that traditional costume has become an emerging and demanding market segment today. The market segment is evident in Sri Lanka's suburbs in Colombo, Kandy, Galle, Rathnapura, and Kurunegala. The researcher predicts that the market meets niche market parameters
item: Conference-Extended-Abstract
Development of an odorless rapid composting machine for domestic use
(Engineering Research Unit, 2023-12) Darshani, AA; Udunuwara, UGPL; Sewwandi, HPG; Madhuwanthi, JAGS; Jayasekara, JGAS; Weerasinghe, L
In the current world, the escalating waste problem has reached an alarming level. With the rapid growth of the population, the amount of waste generated has also increased, leading to significant challenges in its organic decomposition. Like many other regions, Sri Lanka is also facing many challenges due to inadequate waste management practices. One of the areas that help to solve this problem effectively is the development of waste recycling technology. Municipal solid waste takes a long time to decompose naturally and often produces unpleasant odors. Therefore, it is essential to develop an appropriate technology to accelerate their decomposition to manage the problem at the grassroots level. This research aims to provide a technical solution to these pressing issues in waste management and recycling. The scope of this research is to improve an ordinary kitchen garbage composting machine by introducing controlling variables such as temperature and humidity, heating capacity, aeration, and optimization of the local climate within the composting chamber to develop a conducive environment for microbial activities. This would enhance the microbial growth rate and increase the quality of compost while reducing the residence time of garbage. Other than this the research aims to reduce the odor during the recycling.
item: Conference-Extended-Abstract
Design of refuse collection mechanisms for an intelligent beach cleaning robot
(Engineering Research Unit, 2023-12) Yathunanthanasarma, B; Barathraj, M; Mahiliny, J; Jayasekara, AGBP
Ensuring the protection of natural habitats from growing coastal pollution includes the important task of cleaning plastics from beaches. The buildup of plastics and debris not only endangers marine life but also amplifies the problems, underscoring the need for effective intervention measures. While manual beach cleaning efforts are admirable, they are time-consuming and can lead to health issues for workers, including musculoskeletal diseases. Moreover, relying solely on manual labor proves inadequate given the large scale of the problem. Thus, there is an urgent need for advanced technological solutions capable of assisting and enhancing human efforts. This paper advocates for the deployment of intelligent beach cleaning robots as a promising approach towards a more sustainable and efficient strategy for mitigating coastal pollution.
item: Conference-Extended-Abstract
Investigating pomegranate rind dye for wearable health monitoring: a colorimetric approach to analyzing sweat, glucose, and insulin
(Engineering Research Unit, 2023-12) Perera, WTAN; Medagedara, MH; Herath, CN
This study investigates the potential of pomegranate rind dye, a natural source of pH-sensitive anthocyanins, in the field of health monitoring through the colorimetric detection of sweat analytes, markers of physiological states [1], [2]. Addressing the limitations of cost and complexity inherent in traditional sweat analysis methods prevalent in healthcare and sports science [2], [3], our approach utilizes colorimetric analysis. This technique, simpler and more cost-effective, relies on visual color changes in response to specific analytes for quantitative data acquisition [1]. Emphasizing sustainability and economic feasibility, our research highlights the potential of natural dyes, particularly pomegranate rind, in creating eco-friendly, cost-efficient textile-based indicators for health monitoring applications, including fertility and menstrual cycle tracking.
item: Conference-Extended-Abstract
Development of a new bio-based adhesive for cardboard using latex of pterocarpus indicus
(Engineering Research Unit, 2023-12) Dileepa, KDR; Uresh, KA; Udayakumara, SV
Bio-based adhesives are derived from renewable resources such as plant-based materials, animal by-products, and microbial sources. Latex-based bio adhesives, specifically those made from natural latex derived from plant sources, have garnered significant research interest. While natural rubber remains the most commonly used bio source for latex adhesives, this study explores the potential of Pterocarpus indicus wild (PIW) latex as an alternative and discusses modifications that can be made to enhance its adhesive properties. Additionally, this research examines the effect of additives such as Polyvinyl Alcohol (PVA) on the adhesive properties of natural latex while evaluating their capacity for adhesive performance. The major challenges faced by existing bio-based adhesives, including low bonding strength and water resistance are acknowledged. Consequently, this study offers guidance on advancing new plant sources for bio-based adhesives, addressing the limitations of current adhesive products. Adhesives are social substances and can be defined as a mixture in a liquid or semi-liquid state, capable of joining permanently to surfaces, by an adhesive process [1]. Bio- based adhesives are a type of adhesive that is made from natural and renewable resources, such as plant-based materials [2]. These adhesives are becoming increasingly popular due to their sustainability and environmentally friendly nature, as they have a reduced carbon footprint and are biodegradable. The disadvantages of the current bio sealants can restrict their use in some circumstances. Their lower strength and longevity compared to conventional adhesives are significant drawbacks. Some bio-based adhesives are also moisture-sensitive and may lose their adhesion when exposed to moisture or high humidity [3]. Due to the higher expense of obtaining and processing natural materials, bio- based adhesive production has another disadvantage. Additionally, bio-based adhesives' scalability and broad use may be constrained by the lack of readily available natural materials. Additionally restricting their use in some applications, some bio-based adhesives have a lower shelf life than conventional adhesives.
item: Conference-Extended-Abstract
Intelligent wheelchair controller system for human-robot interactions
(Engineering Research Unit, 2023-12) Sarathchandra, HAHY; Priyanayana, KS; Jayasekara, AGBP; Gopura, RARC
Robot controllers are crucial when executing high-level algorithms. Sometimes designing these controllers from scratch will be the best solution for building robots. However, the process will consume much time to build all the electronics and other mechanical hardware from scratch. Therefore, for rapid prototyping continuing from the available hardware will help a lot to save time. Additionally, low-level controllers are crucial when it comes to autonomous robot manipulations [1]. The outcomes produced via these embedded systems [2] will directly pass to the actuators to operate robots. Therefore, the accuracy of these controllers will also cost the overall accuracy of the entire system [3]. There are many types of controlling mechanisms to control wheelchairs. The pieces of evidence from the literature give many examples[4-7]. Out of these controlling techniques, controlling via a joystick [8] is the default controlling mechanism for almost all wheelchairs. When designing novel controlling mechanisms, the way of perceiving environmental details and delivering control signals that are produced with the algorithms is very important and crucial. When exploring the literature, it shows that layered architecture is preferred for implementing controller mechanisms[9,10]. However, the possibility of integrating hardware systems on top of available designs are challenging task to achieve. Even though the literature shows many developments for controlling mechanisms, scaling these hardware units to align with the present knowledge expansions on the algorithms is far below the expected level. Therefore, this work proposed a novel layered hardware unit to replace existing joystick-based controllers and integrated software modules with the help of the proposed layered architecture. Moreover, the work suggests injecting the data by replacing the already available joystick controller of the wheelchairs. In the rest of the article, section II will introduce proposed architectures, and section III will showcase the results based on the observations and follow up with the discussion. Section IV will conclude the paper with future directions.
item: Conference-Extended-Abstract
Experimental bending performance characterization of sHAMs used underwater growing robot
(Engineering Research Unit, 2023-12) Malinda, HAN; Marasingha, MMTM; Senarath, SCD; Dassanayake, PC; Kulasekara, AL
Soft-growing robots represent an emerging field characterized by their ability to extend at the tip while the base remains stationary. These robots offer a unique advantage in navigating through confined spaces inaccessible to humans. Specifically, in underwater exploration tasks, their flexible bodies render them particularly efficient. Moreover, these robots exhibit the capability to steer even while growing. There exist several steering methods including sHAMs (soft Hydraulic Artificial Muscles), sPAMs (soft Pneumatic Artificial Muscles), tendon-driven techniques, and predefined bending mechanisms, applicable both in air and underwater [1],[2]. This paper specifically delves into the study of sHAMs (soft Hydraulic Artificial Muscles) and their inherent characteristics. Various critical factors influence the bending behavior of sHAMs, including pressure, the diameter ratio between the robot body and the sHAM, the length of the robot body, and the number of sHAMs employed. In this study, we concentrate on varying the diameter ratio between the body and sHAM, while keeping other parameters constant. To achieve this, we fabricated robot bodies with varying diameters while maintaining a constant diameter for the sHAM. By systematically analyzing the results obtained from this experimental setup, we aim to provide insights into how the bending angle varies in relation to the diameter ratio.
item: Conference-Extended-Abstract
A Single-feed 3-in-1 cinnamon processing machine
(Engineering Research Unit, 2023-12) Kannangara, KKDVN; Indikadulle, IKKD; Silva, GAR; Gamage, JR
Cinnamon holds global recognition as a necessary spice and a vital food ingredient. Apart from being highly regarded as a spice, it shines in areas such as medicine, as a powerful agent against cancers, heart diseases, and diabetes [1], [2]. Various cinnamon breeds exist worldwide. True Cinnamon, native to Sri Lanka and commonly known as Ceylon Cinnamon, is renowned for its superior taste, aroma, and quality. It contains significantly less coumarin, a potentially harmful substance when compared to alternatives such as Cassia cinnamon [3], [4]. Cinnamon trees yield various extractants, including cinnamon oil from leaves. The most sought-after extract is obtained from the inner bark, which is the primary focus of this project. The cinnamon industry exhibits remarkable resilience and growth, continuing strong export performance during the pandemic. While it faces fierce competition from Chinese Cassia in terms of quantity, Ceylon cinnamon remains unrivalled in quality. Cinnamon extraction is a labour-intensive process cantered around harvesting the prized inner bark of cinnamon stems. Skilled workers are pivotal for ensuring product quality and process efficiency. High labor costs and the need for specialized skills in cinnamon processing pose significant challenges, leading to demotivation among planters and exacerbating the labor shortage issue. In addressing above gaps, this research aims to design and develop an automated machine for cinnamon processing that includes the functions: scraping, rubbing, and peeling in a single feed, as a solution for the high demand of skilled laborers, with high efficiency and acceptable quality output.
item: Conference-Extended-Abstract
A Coconut deshelling machine for improved safety and productivity
(Engineering Research Unit, 2023-12) Bandara, A; Jayawickrama, D; Kuladasa, B; Dinuka, A; Subasinghe, LU; Gamage, JR
Coconut is the third largest cash crop grown in Sri Lanka. There are a variety of coconut-based products available, and deshelling is an essential step in the manufacturing processes of all those products. However, the deshelling process faces several safety concerns which are yet to be addressed. It was found through field visits that on average, 4 accidents per month are reported by the laborers who work in deshelling machines. These injuries are caused by slippage of coconut when it is held onto the cross cutter of the deshelling machine. Another concern was that the coconut kernel was damaged slightly by the cross-cutter, the cutting wheel of the current machine. This damage is significant when a large-scale coconut processing plant with over 200,000 nuts daily capacity is considered. The scarcity of skilled laborers to operate the deshelling machines is another challenge. Training new laborers is both a time-consuming and costly process. Therefore, this research aims to develop a new mechanism for coconut deshelling to be used in an industrial setup. The mechanism intends to enhance the safety of the operation while reducing the damage to the kernel and to reduce human intervention in the process. The scope of this project is limited to deshelling a coconut starting from dehusked coconut and the intended output is kernel with brown skin while dehusking and paring will be out of the scope.
item: Conference-Extended-Abstract
Digitalized platform to better utilize empty space of return runs of truck
(Engineering Research Unit, 2023-12) Bemindu, J; Sharic, S
Sri Lanka depends heavily on road for freight transportation where around 97% of domestic freight is transported by roads using trucks [1]. It was further related that half of the above trucks returning empty [1]. Such empty truck return runs go unnoticed in Sri Lanka. In other developed countries, digitalization such as digital tracking and tracing is used to minimize empty return runs of trucks by making them noticed by the parties in the supply chain. No such digital tracking and tracing is available in Sri Lanka to provide real time information on the truck operations. Therefore, the empty return runs remain unnoticed. Because of its being unnoticed, these empty space in the return runs of trucks are not used optimally. Lack of optimal use of these available space in the return runs of trucks incurs high cost of operation and lowers the net profit of truck operators. High cost and lower profit to truck operators leads to increase in charging high transportation cost from the shippers (local farmers). High transportation cost affects all the parties in the supply chain including the customers, suppliers, transport operators and the country. Therefore, the objective of this study is to propose a digitalization approach of tracking and tracing the trucks that could give real time information on truck operations in Sri Lanka. This digitalization approach can help suppliers, transport operators and regulators to better utilize the empty space available in the return runs of trucks. This approach would pave the way for better customer satisfaction as it could reduce the freight transport costs.
item: Conference-Extended-Abstract
Interactive learning facility for PLCs, and simulations using factory IO
(Engineering Research Unit, 2023-12) Adhikari, M; Kalansooriya, G; Rathnayake, S; Jayasekara, AGBP
To improve efficiency, productivity, and worker safety, industries are implementing automation facilities into their processes to take advantage of the rapid development of technologies [1]. Programmable Logic Controllers (PLCs) are one of the major parts of automatic systems in the industry [2]. Learning PLCs and their functions is one major component in learning industrial automation, and hence laboratories pay much attention to instruct students about PLCs and their functions. On the other hand, there is much research being done on intelligent learning platforms such as ALESK, and BYJU’s learning, where they implement student knowledge assessment methods and adaptive teaching methods [3]. This is one part of an ongoing project working on developing an intelligent learning platform for students to learn PLCs, its decentralized connections and simulations using factory IO. With an attempt to identify the hardware components needed to build a ‘Palletizer’, this work tries, • to develop a simulation in Factory IO for students to learn about PLCs, ladder programming and input/output devices, • to develop a web platform for students to refer to lab materials, data sheets and tutorials. As outcomes of this work, a simulation and the ladder logic program of a palletizer was built, and the components needed to build a physical palletizer and their specifications were identified. Moreover, a web interface was designed for students to learn about the lab practical.
item: Conference-Extended-Abstract
Human centered study of strawberry picking behavior for intelligent harvesting robot
(Engineering Research Unit, 2023-12) Sendilkumaran, S; Pragalathanan, A; Laksman, P; Jayasekara, AGBP
Strawberries, highly valued both commercially and for consumption, are predominantly cultivated in controlled environments like greenhouses. Despite these advantages, human labor remains a significant factor in strawberry production costs, especially during harvesting. To address this, there is a growing interest in developing strawberry harvesting robots. However, detecting strawberries poses challenges due to leaf shading, overlapping, and plant structure interference. Enhancing the visual object detection system is crucial for accurate and real-time detection. Given the vulnerability of strawberries to damage due to low mechanical strength during harvesting, we propose a stem-based harvesting method and plan to develop an efficient motion planning approach for the robot manipulator [1][2]. The goal is to enable the intelligent robot to navigate autonomously, picking strawberries with improved efficiency and minimal damage, thus transforming strawberry harvesting processes.
item: Conference-Extended-Abstract
Analysis of cyber attacks in power grids with increasing renewable energy penetration
(Engineering Research Unit, 2023-12) Dayarathne, MASP; Jayathilaka, MSM; Bandara, RMVA
With the application of advanced computer and communication technologies, traditional power systems are converting to cyber-physical power systems. [1] Several secondary systems, like SCADA (Supervisory Control and Data Acquisition), WAMS (Wide Area Monitoring System), AMI (Advanced Metering Infrastructure), and smart substations, are the main cyber-physical subsystems that could be vulnerable to cyber-attacks. [1], [3] Mathematically modeled cyber-attack types such as FDIA (False Data Injection Attack) [1], [2], [3], [4], DoS (Denial of Service) [1], [3], replay attacks, etc. can be modeled using the PSCAD power system model with the addition of renewable power source models to the system. The goal of this project is to develop a PSCAD power system model with increasing renewable power sources, simulate the different kinds of cyber-attacks mentioned above in the PSCAD power system model, collect faulty data, and develop a deep learning model to detect and act against the cyber-attacks.
item: Conference-Extended-Abstract
Bimanual tele rehabilitation robot
(Engineering Research Unit, 2023-12) Perera, KLM; Hettiarachchi, TI; Kumara, JAAS; Ruwanthika, RMM; Arunya, PSSD
Post-stroke patients lose their interlimb coordination and ability to do bimanual activities. This is common for arms, and usually, patients get one arm paretic. This is a heavy burden since everyday activities need both hands to do. For recovery, they need intensive bimanual therapy. The robotic approach is more effective than traditional rehabilitation for this kind of therapy. A bimanual master-slave robotic system can be used to do simultaneous exercises for patients’ both hands. Our focus is to implement the two most basic and critical movements into this robot. One is Flexion and Extension, and the other is Internal and external rotation. Stroke is one of the conditions that causes long-term disabilities. This causes a lot of struggles for patients to recover. To overcome that and speed up the recovery we propose a robotic approach, with reduced dependency on a therapist and increased repeatability and robustness. When doing these exercises, the patient’s shoulder must be in a relaxed position. Otherwise, the generated torques will depend on the shoulder position. Hence there must be adjustments for varying arm lengths as it changes the position of the elbow joint. Also, we are going to implement two exercises which are done in perpendicular planes into one 1 DOF (Degrees of Freedom), robot. And we need to provide methods to change the motor orientation for that.
item: Conference-Extended-Abstract
Automated guided vehicle for carrying carts
(Engineering Research Unit, 2023-12) Kirushan, T; Punsala, PPGP; Wimalajeewa, MHTN; Jayasekara, AGBP
The growing demand for efficient and reliable transportation solutions in industrial and commercial settings is driving the development of innovative technologies, such as automated guided vehicles (AGVs). AGVs are unmanned vehicles that can navigate autonomously along predefined paths, offering a number of advantages over traditional transportation methods, including increased efficiency, reduced labor costs, and improved safety. However, conventional AGVs are typically designed to transport specific types of materials, limiting their versatility in industries such as food and beverage, where both solid and liquid materials need to be transported. To address this challenge, we propose an AGV design capable of transporting both solid and liquid materials. It utilizes a line-following navigation system with a load type identification system and a speed controlling system, to ensure safe and efficient transportation of both solid and liquid materials. In this paper, we present a detailed overview of our AGV system design and implementation, highlighting its key features.
item: Conference-Extended-Abstract
Neural network based model for estimating the resistance of outdoor distribution substation grounding
(Engineering Research Unit, 2023-12) Madhushan, PN; Pabasara, WMN; Shihab, M; Gunawardana, M
Grounding is one of the most important parts of an electrical system. Earthing systems are done to protect the power system and the personnel from the danger of electrical shocks. Ceylon Electricity Board (CEB) uses a special structure for transformer earthing arrangement. The used structure is copper bonded earth rod with a concrete filled steel cage. Due to the complexity of the structure and the nonlinear variations in soil parameters, it is challenging to determine resistance before implementing the structure. We can use an analytical formula for structures to find the resistance. [4] But for complex structure, as we use here, it is challenging to produce an analytical formula. The other solution is to use a Finite Element Method(FEM) to solve the problem.[2] But is also a time-consuming task.[1] So, we propose a combination of FEM and a neural network-based solution for this task.[1]. We propose to generate a data set using FEM and implement it in a neural network.
item: Conference-Extended-Abstract
Deep learning-based power baseline modelling of a range of electrical loads in smart green buildings
(Engineering Research Unit, 2023-12) Gunawardhana, KVSD; Lakshitha, WHAS; Perera, ULDE
Energy consumption modelling of electrical loads plays a crucial role in modern-day energy management systems for green commercial buildings. This project focuses on the development of power baseline modelling using deep learning techniques for a diverse range of electrical loads. The baseline model, an estimation of power or energy consumption before implementing energy management, is widely used to identify savings by comparing with the measured data after implementing energy management. Energy efficiency is crucial for both commercial and non-commercial applications. Power baseline models give reference to identify energy saving or loss. We can assess energy saving after implementing energy conservation strategies and identify energy wastage of the system when actual power consumption is higher than the power baseline model prediction. In this study specifically, a comparison is made between Karl's Pearson's and Random Forest-based deep learning approaches and Recurrent Neural Network (RNN) models. This project incorporates both simulations and real-world data to conduct the study.
item: Conference-Extended-Abstract
Recycling plastic waste for enabling circular economy
(Engineering Research Unit, 2023-12) Haameem, H; Aabdeen, F; Sharic, S
Municipal solid waste management is a major issue for many nations striving for a sustainable future. According to the United Nations Environmental Programme report [1], Municipal solid waste management is both a resource and a burden as it contains valuables and leftover waste that need to be managed carefully to keep the public healthy. Biodegradable and non-biodegradable components make up the majority of this waste; the latter includes things like plastics, tires, e-waste, glassware, building debris, metals, ceramics, some textiles, and batteries. Notably, there is significant worry about non-biodegradable waste, especially plastics. Sri Lanka is currently experiencing an economic crisis, import restrictions, shortages of goods, and hike in foreign exchange rate, all of which highlight the need to adopt a circular economy. Natural ecosystems are seriously threatened by Sri Lanka's rapidly increasing plastic usage, which is increasing at a pace of 16% per year [2]. Every year, over 265,000 megagrams (Mg) of plastic are consumed [2], most of which wind up in landfills. Sri Lanka has not yet adopted practices akin to those of other nations that recycle plastic waste into the circular economy. Failure to convert plastic waste could lead to adverse impacts such as reliance on imports, financial losses, unstable foreign exchange rates, shortages of necessities, rising demand for plastics, problems with the carbon cycle and environmental equilibrium, resource depletion, and failure to meet the Sustainable Development Goals.
item: Conference-Extended-Abstract
A Real-time, scalable and extensible object filtering and detection system using kinect sensor and ROS2 foxy
(Engineering Research Unit, 2023-12) Gunasekara, CM
Object detection and filtering based on shape and color is an important capability for many robotics applications. For example, sorting objects by shape and color is a common industrial application. Service robots also need to detect and track objects based on visual properties. While powerful deep learning approaches like YOLO have emerged for general object detection, they require large datasets and extensive training. A simple shape and color filtering provide a lightweight and customizable alternative. This work aims to provide a real-time modular and lightweight system that can identify objects of basic shapes and colors and allow extensibility of functionality by incorporating more custom color and shape filters. The proposed system for real-time shape and color filtering using a Microsoft Kinect RGB-D sensor and Robot Operating System (ROS2) can identify an array of regular shapes like circles, rectangles, and triangles over a spectrum of different colors. New shape and color filters can be added dynamically at runtime thanks to the modular, ROS-2-based implementation.

Browse

Recent Submissions