Purpose: This paper aims to develop a coffee supply chain traceability system based on Blockchain (BC) and Machine Learning (ML) with the aim of ensuring the quality of coffee beans production. BC functions to ensure supply chain performance, while the ML model ensures product quality.Design/methodology/approach: Smart Contracts will be built on the Ethereum Virtual Machine BC network based on Ethereum. The ML model to identify good and bad green coffee beans will be built using different YOLO algorithms, which will go through training and validation stages, namely using the k-fold cross validation method. The ML model algorithm is based on Convolutional Neural Network (CNN) using YOLOv5m, YOLOv6m and YOLOv7. The best model will be chosen based on the results of cross-validation with test data in the form of coffee image data that the model has never seen (unseen data). The whole process of building the ML model is done on the Google Collab Pro+ Virtual Machine.Findings: YOLOv5m outperformed the other models in both non-augmented and augmented training datasets, highlighting the proficiency of YOLOv5m in managing compact datasets and its resilience in the face of data augmentation, positioning it as a prime selection for quality discernment tasks within the realm of green coffee beans. The smart contracts offer an all-encompassing approach for user management, monitoring product status, and presenting traceability data within the framework of coffee plantation administration.Originality/value: This research contributes to the development of blockchain network as a solution to implement traceability systems along coffee supply chains in Indonesia. Moreover, it shows that while blockchain can ensure the process of production along the coffee supply to follow certain guidelines, machine learning can verify whether the product that was produced by utilizing BC is of high quality/acceptable.

Authentication is needed to enhance and protect the system from vulnerabilities or weaknesses of the system. There are still many weaknesses in the use of traditional authentication methods such as PINs or passwords, such as being hacked. New methods such as system biometrics are used to deal with this problem. Biometric characteristics using retinal identification are unique and difficult to manipulate compared to other biometric characteristics such as iris or fingerprints because they are located behind the human eye thus they are difficult to reach by normal human vision. This study uses the characteristics of the retinal fundus image blood vessels that have been segmented for its features. The dataset used is sourced from the DRIVE dataset. The preprocessing stage is used to extract its features to produce an image of retinal blood vessel segmentation. The image resulting from the segmentation is carried out with a two-dimensional image transformation such as the process of rotation, enlargement, shifting, cutting, and reversing to increase the quantity of the sample of the retinal blood vessel segmentation image. The results of the image transformation resulted in 189 images divided with the details of the ratio of 80 % or 151 images as training data and 20 % or 38 images as validation data. The process of forming this research model uses the Convolutional Neural Network method. The model built during the training consists of 10 iterations and produces a model accuracy value of 98 %. The results of the model's accuracy value are used for the process of identifying individual retinas in the retinal biometric system.

The analysis of the performance of higher education academic quality in terms of student achievement, study period, and drop out rates is still an intensive study among researchers. Several clustering methods are often used to understand student and graduate groups, in influencing college performance. However, the conventional method has only arrived at the results of clustering, so it is difficult to interpret it as a support for academic decisions, especially in mapping the position of universities against other universities nationally. This article introduces a combination of techniques from self organizing map and technique for oorder preference by similarity to an ideal solution (SOM-SIS), an auto-summarizing technique from clustering results as well as mapping university academic performance. First, the academic performance indicators are grouped using the self organizing map (SOM) method and the results are concluded using the technique for order preference by similarity to an ideal solution (TOPSIS) approach. The SOM-SIS technique was tested using data from one of the universities in Indonesia. As a result, the SOM-SIS technique has a 100% compatibility rate with the higher education quality assurance system, through recommendations from three university experts.

Indonesia is one country that has enormous potential in the use of medicinal plants as herbal medicines. Utilization or use of plants, especially medicinal plants as a means of healing disease has long been used. However, people in general still have difficulty knowing the types of plants that can be used as herbal medicines. This is due to the limited information and knowledge possessed by the community to identify and identify the use of medicinal plants. This study describes the development of feature extraction in leaf images for the identification of medicinal plants, where the main difficulty in the leaf identification stage is that the morphological (physical leaf shape) and physiological (leaf shape characteristics) are different for each type of leaf. There are three methods proposed in this research, namely the first is the proposed leaf feature model in the form of 16 perimeter point distances to leaf centroid points and seven median line connectors. The second is to develop leaf feature extraction methods and algorithms so that 23 leaf shape features can be generated for each type of medicinal plant. Third, making a prototype identification system or the introduction of medicinal plants based on leaf morphological characteristics. The identification process is carried out using two approaches, namely the Manhattan Distance and Artificial Neural Networksimilar. In the testing phase of the resulting software prototype, 51 types of medicinal plant leaves were used where each type consisted of 10 different leaf images. Based on the trial results, the accuracy rate of identification or recognition of medicinal plants using Manhattan Distance is 99.0196%, and when using Neural Networks, the accuracy rate reaches 100% for training data and 84.31% for testing data.

Road pavement is a supporting factor for national development, especially in the distribution of trade in goods and services as well as the movement of human mobility. Road maintenance needs to be done regularly so that the road is always in good condition, but the weather and road loads are the things that cause road damage. Road damage is generally categorized into cracks, alligator cracks and potholes. The purpose of this research is to utilize image processing to detect and classify the types of road damage. The steps involved include: image acquisition with a digital camera, conversion of RGB images into grayscale images, image normalization, selection of damage points, counting histogram bins, determining damage bins, calculating noise with image morphology (closing and opening) using a disk element structure of size 5, calculating radial vector and finally classifying road damage using the K-NN (K Nearest Neighbor) method with 3 classes and a K value of 11. The image from the classification results is then calculated the level of damage based on the category according to the SDI (Surface Distress Index) provisions, where the level of crack damage is seen from the width of the crack, the alligator crack is seen from the percentage of damaged area compared to the segment under review and the pathole is seen from many holes. The test used 597 images consisting of 95% training data and 5% test data, the results obtained that the accuracy of this research reached 83%.

Waste-to-energy (WTE) is one of the Indonesian government’s programs aiming to meet the target of achieving a new and renewable energy (NRE) mix, as well as one of the solutions proposed to overcome the problem of waste. One of the products of WTE is energy derived from raw material waste (refuse-derived fuel/RDF). Using the formula y = 0.00003 x⁵ − 0.0069 x⁴ + 0.6298 x³ − 24.3245 x² + 432.8401 x + 55.7448 with R² = 0.9963, which was obtained by comparing a scatter plot diagram from the RDF calorie test dataset produced through a bio-drying process, the potential RDF calories produced using the waste composition dataset taken from each region in Indonesia can be calculated. The results of the calculations using the determined equations produce a list of provinces with RDF calorie potential, ordered from the largest to the smallest, using which the government can determine which areas are the main priority for processing waste into energy. Thus, through this method, the target of 5.1% renewable energy sourced from waste can be achieved by 2025.