Integrating Internet of Things (IoT) devices and machine learning (ML) techniques holds immense potential for transforming beekeeping practices. This review paper offers a critical analysis of state-of-the-art IoT-enabled precision beekeeping systems. It examines the diverse sensor technologies deployed for honeybee data acquisition, delving into their strengths and limitations, particularly regarding accuracy, reliability, energy sustainability, transmission range, feasibility, and scalability. Furthermore, this paper dissects prevalent ML models used for bee behaviour analysis, disease detection, and colony monitoring tasks. This paper evaluates their methodologies, performance metrics, and the challenges involved in selecting appropriate machine learning algorithms. It also examines the influence of sensing devices, computational complexity, dataset limitations, validation procedures, evaluation metrics, and the effects of pre-processing techniques on these models’ outcomes. Building upon this analysis, this paper identifies key research gaps and proposes promising avenues for future investigation. The focus is on the synergistic use of IoT and ML to address colony health management challenges and the overall sustainability of the beekeeping industry.

The changing environment, climate, and the increasing manifestation of disasters, has generated an increased demand for accurate and timely weather information. This information is provided by the National meteorological authorities (NMAs) through different dissemination channels e.g., using radios, Televisions, emails among others. The use of ICTs to provide weather information is recently gaining popularity. A study was conducted in three countries, namely Nigeria, Uganda, and South Sudan to assess the efficiency of an ICT tool, known as “Weather Information Dissemination System”. The study involved 254 participants (Uganda: 71; South Sudan: 133; and Nigeria: 50). The collected primary data were first quality controlled and organized thematically for detailed analysis. Descriptive statistics was used to provide quantitative analysis as well as content scrutinized for qualitative analysis. The results showed that there is a need for timely weather information to plan farming activities such as planting and application of fertilizers and pesticides as well as to manage flood and drought by the water sector and disaster management. Results further showed that the majority of the respondents have access to the technology needed to access weather and climate information. The respondents who received weather information from NMAs noted that the forecast was good. However, they further noted that there is more room for improvement especially with making the forecasts location-specific, ensuring mobile access is adequate in all regions, provision of weather information by SMS (in countries where this service is currently unavailable) and improved timing of the weather information. Finally, uncertainty about the accuracy of weather information and the weather information not meeting specific needs are key barriers to people’s willingness to pay for it (Uganda: 33.3%; South Sudan: 46.1%; and Nigeria: 33.3%). Improved collaborations between the NMAs, ICT service providers, policymakers and farmers will facilitate an effective approach to weather information access and dissemination. Innovative sensitization approaches through the media houses will enable better understanding of weather products and utilization, and access to enabling ICTs would increase access to weather forecasts

The changing climate has negatively impacted food systems by affecting rainfall patterns and leading to drought, flooding, and higher temperatures which reduce food production. This study examined the ability of communities to cope with food insecurity due to the changing climate in the Serere and Buyende districts, which are two different agro-ecological zones of Uganda. We administered 806 questionnaires to households, a sample size which was determined using Yamane’s formula, with the snowball sampling method used to select the households. The questionnaire sought information, including that regarding the respondents’ resources, the effects of climate change on households, and the coping mechanisms employed to reduce the impact of climate change on food security. The data collected was coded and analyzed using the statistical package for the social sciences (SPSS). Agriculture was found to be the main source of income for 42.4% of male adults and 41.2% of female adults in Serere. In Buyende, 39.9% of males and 33.7% of females rely on selling animal, poultry, and food crops. Aggregate results further showed that 58.3% of females and 42.2% of the males from both districts had suffered from the impacts of climate change, and that the effects were more evident between March and May, when communities experienced crop failure. The study further found that the percentage of households who had three meals a day was reduced from 59.7% to 43.6%, while the number of households with no major meals a day increased from 1.3% to 1.6%. We also found that 34.3% of households reported buying food during periods of crop failure or food scarcity. Moreover, despite reporting an understanding of several coping mechanisms, many households were limited in their ability to implement the coping mechanisms by their low incomes. This reinforced their reliance on affordable mechanisms, such as growing drought-resistant crops (32.7%), rearing drought-resistant livestock breeds (26.1%), and reducing the number of meals a day (14.5%), which are mechanisms that are insufficient for solving all the climate-related food insecurity challenges. We recommend that the government intervenes by revising policies which help farmers cope with the negative effects of climate change, promoting the sensitization of farmers to employing the coping mechanisms, and subsidizing agricultural inputs, such as resistant varieties of crops, for all to afford.

Unmanned Aerial Vehicles (UAVs), also known as drones, have found applications in precision agriculture owing to their agility and remote access. However, because of battery, bandwidth and environmental constraints, UAV’s networks experience intermittent links and packet drops. These constraints can either force the UAV to abort its mission or increase packet drops. Therefore, there is a need to improve routing protocols and UAV path planning. This work reviewed routing protocols such as reactive and hybrid as well as mobility predictions such as regressive and Markov models with the central argument of topology control and path optimizations. Optimizations minimize the coverage time and energy consumption. As routing protocols operate on the network layer, this review analyses articles that include cross-layer issues, modelling, and predictions. This article also examined QoS metrics (such as bandwidth and link quality), antenna selection and placement on the UAV and the methods used to validate routing protocols. In this work, UAV’s path optimizations consider crop locations, the appropriate placement of ground sensor nodes and how neighbouring UAVs are controlled for efficient data collection and relay. Analysis of the collected data can improve crop health, cultivation management, livestock monitoring, and reduction on operational cost. Therefore, this work resolved that by predicting the topology change, link-lifetime, and residual energy, the routing protocol can reduce retransmission and packet drops, thereby enhancing network longevity.

We propose a prefetching mechanism at the iSCSI initiator to improve the memory cache hit rate and attain better bandwidth utilization. By monitoring the scheduler run queue, the iSCSI initiator is able to move pages to the iSCSI initiator memory cache from the iSCSI target before a page fault occurs. To increase the effectiveness of prefetching, a small striped disk is used to keep more copies of the pages brought in from the iSCSI target. A prefetch analyzer analyzes both page fault and prefetch information which is later used by the iSCSI prefetcher to make better intelligent prefetching decisions. Our algorithm shows a 15% increase in sequential file reads and little or no performance improvement in random accesses.