| dc.description.abstract | As the number of students joining higher institutions of learning increases per year, the facilities and infrastructure of the institutions aren't increasing meaning that they tend to be overwhelmed. This brings about congestion in the available space, collisions and confusion of all the various stakeholders. Some Universities tend to just cut on the number of students they admit to try and curb this problem.
Makerere University got two new structures, Central Teaching Facilities 1 and 2 whose major goal was to find a solution to the underlying problem. These structures ended up falling victim of the same prior problem where rooms were still overcrowded, collisions and some colleges assuming ownership. Because they lacked a proper efficient scheduling system which is strict, optimal and organized.
This created the need for an Automated Scheduling and Room Allocation algorithm to stamp the new problems and more. All university courses would have equal opportunity to use the facilities, rooms would be allocated according to size and special requests for rooms would be handled professionally. An embedded system powered display would be available to always guide the students on rooms, lectures and times they are available or take place respectively. A web-based application interface would provide the people in charge at the facility with remote access of the scheduling algorithm.
To achieve this, we carried out surveys to understand the specific problems faced and how best the stakeholders would want the system to work. After extensive research, the constraint optimization approach was used. This approach enables us break the problem down into arbitrary constraints defined on a set of variables and identify feasible solutions out of a very large set of possible candidates, after which an optimal solution can be acquired. For the web-based application, JavaScript-based frameworks were used that is VueJS, ExpressJS and NodeJS because of their neat structure and overall efficiency. For the embedded systems a raspberry pi was used because of specifically its lower power rating and consumption.
An optimal solution which is the schedule of the lectures was achieved efficiently, that is, earliest time possible and this justified the use of the solver we employed since it outperformed all the others in terms of time even as the number of variables increased. An easy to use web application was also developed and this makes it possible for an administrator to schedule the lectures, and this is displayed using a raspberry pi whose performance also justifies our choice. | en_US |
