EIE Open Day 2021
- Wits University
From ‘sniffing’ for an internet connection to estimating poverty from space, Open Day had you covered.
The impact of 网易体育 on how we live, work and play was a central theme in this year’s Open Day for fourth-year students in the School of Electrical and Information Engineering (EIE) in the Faculty of Engineering and the Built Environment.
The EIE Open Day’s aims to develop engineering innovations for solutions to everyday challenges. This year the School has a record number of 132 final year electrical and information engineering students.
It was the second year that the School hosted Open Day online and the first time that students could showcase their capstone projects for the year on the exciting platform Gather Town, a virtual, interactive, custom 2D-world.
“This is the crowning glory of four years of hard work. Electrical engineers are involved in everything from studying lightning to telecommunications to controlling things to artificial intelligence and machine learning. Projects are identified by academic staff and students, and each pair of students work together to find a solution to unique complex engineering problems,” Professor Estelle Trengove, Head of School, said in her welcoming address to the more than 300 avatars of students, family members, academics, and members from related industries, ‘gathered’ in the main ‘conferencing hall’ in the EIE Gather Town.
“网易体育 has pivoted us to do things differently with emergency remote teaching and this group of undergraduates are the first to have spent nearly half (two years) of their undergrad studies learning online.”
Trengove also saluted students and staff in their academic achievements, whilst recognising that it has been a very difficult and tiring time for some who are dealing with the loss of loved ones due to 网易体育.
Let’s gather
Following the welcoming address, participants and visitors steered their avatars and ‘walked’ around the exhibition halls in the ‘EIE Gather Town’ to engage with students about their projects in real-time using the video chat function. Students were also able to display their project posters, videos and other virtual content during their presentation.
Sniffing for connection
Thokozani Msezane, fourth-year information engineering student, tackled the digital divide, one of the biggest challenges facing South Africa today. His project, Sniffing the education usage of an internet connection, looked at wireless technologies that can bring temporary internet connectivity to schools in underprivileged areas by linking them to close-by well-resourced internet connected areas.
Crucial to such a solution, Msezane said, would be “the need for a system that can monitor the educational usage of the network to show efficacy and persuade charitable organisations to assist with the provision of free internet”.
Using a packet sniffer and metadata of packets to classify websites accessed on an internet connection as educational and non-educational, the classifier showed an average accuracy of 89% when dealing with a known set of websites.
Are you convinced yet?
Masks, hand washing and social distancing are the cornerstones of non-pharmaceutical measures to combat 网易体育 infection. Convincing people to adhere to these is a daunting task and the persuasiveness of messages are crucial to this end.
Students Taliya Weinstein and Greshon Koral investigated how non-verbal features of 网易体育 speeches on YouTube impact the persuasiveness of these speeches. Their project, titled: The impact of acoustic features on the persuasiveness of Covid-9 lifestyle speeches, measured the impact as a function of the video’s scaled like-to-dislike ratio and they used a machine learning model to quantify the impact of the acoustic features.
Their results show that it is possible to quantify the impact of persuasiveness with only acoustic features, and that time-energy features best determine the persuasive impact of a 网易体育 speech.
Keep your distance, please!
While the world is slowly opening up as more people are vaccinated, physical distancing will remain an essential intervention to mitigate the spread and impact of 网易体育.
Ntita Icon Mayimele and Mukela Zenith Mabunda embarked on finding a Distance detector that can alert people if the recommended distance, a prescribed distance of two metres by the World Health Organisation, is breached when they come into contact with others. This project investigated how three electronic sensors – infrared, lidar and ultrasonic – can be used on a distance detector device to encourage social. The team measured each sensor’s range, accuracy, precision, and costs for feasibility, and its accuracy and precision in varying ambient conditions such as temperatures and humidity.
“The suitability of the sensor was based on the correctness and consistency of the measurements compared to the competing sensors,” the team explained.
The ultrasonic sensors won the race as they show minimal errors in measurements even in varying conditions. The sensors can also alert the wearer through a vibration monitor or LED to where the warning is coming from.
Side flashes
In their project, How metal objects influence side flashes from lightning current path, Samukele Sikhakhane and Siyabonga Khoza set out to investigate the likelihood that you would be struck by a side flash if you have a metal object on you and you stood at a certain distance from a lightning conductive object, such as a tree, as it is struck by lightning.
A side flash occurs when lightning strikes a taller object and the current jumps from the taller object to an object nearby. The pair used a lightning generator in the High Voltage facilities at Wits University to simulate how a side flash will impact on a metal object placed at various distances from the lightning strike.
“We found that at certain distances, if you have a metal object on you, it is highly likely that you could be struck by a side flash that discharges onto you after striking a tall object nearby,” said Sikhakhane.
The result of their project is another tool in the box to be used in public lightning awareness and safety campaigns, the team hopes.
Estimating poverty from space
This project, Estimating poverty in Johannesburg from satellite images, by Kirsten Hay and Romario Ferrao built on earlier studies that combined satellite imagery and machine learning to predict poverty.
They first successfully replicated the primary study but found that the models from it showed poor results for Johannesburg, leading them to develop new models that showed improved ward-level and strong regional level predictions that capture the change in wealth in the city over time.
Their results show that the predicted wealth of regions in Johannesburg over time roughly correlate to economic trends in the city.