Urban farming has the potential to utilise unused space in the community to alleviate food shortages and increase the community’s income through local food production. When Internet of Things (IoT) technology is integrated with urban farming, it can further improve its efficiencies and yield. The work in this paper improved our previous work of using an IoT-based climate control system to regulate the cultivation environment of oyster mushrooms automatically. Even though the climate control system could produce two batches of mushroom yields, there were several limitations, such as less efficient climate control due to threshold-based corrective action, water wastage, and system instability. This paper aims to address these stated limitations by implementing a fuzzy logic algorithm and redesigned the climate control system. Two crisp input variables from DHT22 sensors representing temperature and humidity were fed into the Node MCU microcontroller’s fuzzy logic coded in C language. The temperature and humidity conditions were divided into five fuzzy trapezoidal membership functions resulting in 25 fuzzy rules to control the duration of running the water pump and ventilation fan. An internal, lightweight web server were managed all HTTP client requests. The enhanced system also included a safety measurement to avoid overheating the microcontroller and causing water wastage. Upon analysis of the data captured in two months, the result showed a decrease of 40% in water utilisation and an increase of mushrooms yield up to 226%. The enhanced climate control system also facilitated maintaining and controlling the temperature and humidity conducive for optimal mushroom cultivation.

No abstract

Abstract-Project-based learning (PBL) is a teaching method that integrates real-world problems and technology into the curriculum. The use of mobile and video technologies in PBL has made the learning environment much more interactive. However, the gap between the richness of video content and the limited capabilities of mobile devices still remains a challenge. In order to solve this problem, video transcoding can be implemented. However, this solution may provide extreme CPU consumption and processing delay on the server's side. To overcome this issue, a private cloud-based video transcoding architecture is proposed. In this paper, the proposed solution was designed, implemented and compared with the server-based content switching method, and an experiment was conducted to evaluate the method in large-scale simulated workloads. The results show that the proposed architecture is capable of dealing with many concurrent users' access and capable of reducing the video downloading time.Index Terms-Cloud computing, project-based learning, video transcoding.

Severe uncertainties climate changes course flood and droughts disaster have made clean water precious for domestic consumption. Thus, securing clean water is important. Wastage of water comes from water consumption such as from household usage. However, monitoring water consumption from household usage is tedious and time consuming. This work utilized Genetic Algorithm (GA) to optimize the coefficient of micro-components of water consumption (CMWC) values to determine high influential household routine parameters. Nine household parameters have been investigated namely, bath/shower, personal hygiene, flush toilet, wash cloth by hand, wash cloth by washing machine, food preparation, water plant, washing car and miscellaneous. These parameters are encoded as a chromosome data in GA to incorporate the CMWC values. The aim is to minimize the residential water consumption estimation error rates and subsequently enabling increased accuracy towards estimating and classifying the amount of residential water consumption. Data average monthly water consumption were collected from 80 households in Seremban. Water consumption has been categorized into three groups of low (L-PDWC), medium (M-PDWC) and high (H-PDWC). Comparison was made between per capita water consumption (PCC) and Domestic Water Consumption via Genetic Algorithm (DWC-GA) error rate’s values. The results are as follows; PCC method’s error rates of 9.49 and DWC-GA error rate is 1.05.