The Department of Electrical and Electronic Engineering, specialises in the control and automation of space satellites, aircraft, and ground vehicles for many different applications. The Electronic Systems Laboratory (ESL) was founded in 1992 and developed, amongst others, SUNSAT, the first South African satellite launched and commissioned in 1999. Since 2004, the ESL department has cultivated a strong research focus on robotics and autonomous vehicles aiming to solve real-life problems. This resulted in the creation of a number of private companies. One of the ESL’s current areas of research is autonomous systems including autonomous vehicles, drones and so forth. This entails developing vehicles that can function completely independently, and are self-directed, self-governed, and self-sufficient. In true South African style, practical research is conducted with the aim of solving real control system problems. Practical research in this context means testing the vehicles’ self-sufficiency when presented with real-life conditions such as weather variability or dirt roads with many obstacles and not only in a controlled laboratory environment. The result: Independent robot-vehicles having to determine – without a map or clearly-defined road – their location, the terrain they are operating on and how to navigate to the endpoint, autonomously. Nevertheless, it is evident that autonomous vehicles have the potential for extremely wide fields of application for example when -used in mining operations these vehicles have the potential to be completely self-sufficient; or, when used in search and rescue operations it can be more effective and safe using autonomous vehicles. How do we create autonomous systems that can handle real-life issues? This will be the subject of our next post.

Internationally renowned scientists, Professors Leon Dicks and Willie Perold, have joined forces to create a proof-of-concept nanowire biological sensor (or nanochip) that can identify any pathogen or bacteria, for example, e. coli, Salmonella or cholera in 10 to 15 minutes after being swallowed. The significance of this research is that it can ultimately be utilised to accurately diagnose patients during an epidemic or outbreak using a combination of nanotechnology and microbiology. It also means that “instead of prescribing a broad-based antibiotic, or waiting 48 hours for the lab tests to come back, a doctor will be able to immediately prescribe the correct antibiotic to target the pathogen, and by doing so, put less stress on the body’s immune system.”[1] In layman’s terms it means soon you will be able to swallow a nanochip that can diagnose a bacterial infection faster and more cost-effective than traditional tests including bloodwork and endoscopies. Furthermore, you would be able to isolate specific bacteria to test for. For example, if your concern is water quality, you can set up the sensor to only look for e. coli. Background During a discussion between Prof Perold of the Department of Electrical and Electronic Engineering and Prof Dicks of the Department of Science, the two scientists realised that they could join forces and combine their research to develop something unique to meet a specific need in the medical world. At the time, Prof Dicks was developing a gel capsule, which if swallowed, had the potential to diagnose whether a patient had a certain disease or infection. The capsule contained antibodies, and once the capsule was removed it could be established whether the antibodies bonded with the pathogen on the capsule which was the indicator for the presence (or not) of a certain disease or infection. At the same time, Prof Perold’s PhD student, Stanley van den Heever, was working on a nano zinc oxide generator that could generate electricity the moment the nanowires are disturbed.  Together they came up with the idea to “combine the sensor with ‘biological bait’ to selectively attract bacteria”. The movement would then generate an electrical signal indicating the presence of the selected bacteria. How it works To prepare the nanogenerator for biological use, a silicon chip of 1cm by 1cm had to be constructed. Zinc oxide molecules are subsequently stacked on top of each other to form a nanowire that is only visible with an electron microscope. Thousands of these nanowires are positioned in such a way that the slightest disturbance of their structure will lead to, what is known as, piezoelectric energy. This energy is then converted to electrical energy and amplified to produce a voltage reading.   After this process, the microbiologists tested the concept by attaching lysozyme molecules (small disease-fighting proteins found in our saliva) to the tip of each nanowire. As soon as antibodies specific to the lysozyme react to the nanowires, it resulted in an alignment shift of the zinc oxide molecules. This caused a change in electrical output which proved that the concept worked.     If you reverse engineer this concept, it is also possible to attach specific antibodies to the nanowires, which will detect the antigens characteristic of a specific pathogen (that is, a disease-causing microorganism) and report the pathogen presence within seconds. Prof Dicks indicated that the “key to the concept is linking the correct antibody to the nanowires to form a perfect, “one and only”, match with the antigen”.[2]     The future According to Prof Dicks, their ultimate hope is to transfer the electrical signal explained above, which is selected to be unique to each pathogen, to a receiver such as a smartphone. This means that 1) you can enter remote areas where epidemics or outbreaks are likely to occur and diagnose the disease in minutes as long as you have antibodies available; and 2) it  is a more cost effective exercise than, for instance, a endoscopy that requires instruments, a venue and a specialist. Whether the concept will also be able to identify viruses like Ebola, Prof Dicks is convinced that it will be able to do so as long as you have antibodies specific to the unique Ebola virus antigens. The latter are usually “proteins located on the surface of cells. These proteins act as a ‘signature’ of that specific cell. This ‘signature’ is then recognised by specific antibodies.”[3] This will be a massive step forward in the early diagnosis of Ebola as, traditionally, Ebola is hard to diagnose due to the similarities in its early symptoms to diseases such as malaria, typhoid fever, shigellosis, cholera, leptospirosis and meningitis. [1] https://blogs.sun.ac.za/microbiology/2014/08/21/nano-chip-in-a-capsule-identifies-bacterial-infections-within-minutes/ [2] https://www.popularmechanics.co.za/science/scientists-develop-sensor-to-identify-disease-causing-bacteria-in-minutes/. [3] https://www.popularmechanics.co.za/science/scientists-develop-sensor-to-identify-disease-causing-bacteria-in-minutes/.

The rhinoceros has few natural predators. Yet, as a result of active poaching, the African black rhinoceros (Diceros bicornis) and the African white rhinoceros (Ceratotherium simum) are nearing extinction.[1] South Africa is home to approximately 80% of the world’s rhinos and as a result, has been hardest hit. Recent poaching statistics released by the South African Department of Environmental Affairs indicated that the period 2007 to 2014 saw an increase of 9 000% in the number of rhinos killed for their horns, with more than half of these poaching incidents occurring in the Kruger National Park.[2] Despite the numbers slightly decreasing for the period 2015 to date, “two and a half rhinos are still killed every single day”.[3] Poaching stats Background: South African rhino poaching The driving force behind the illegal killing of rhinoceros is the relentless demand for rhino horn from Asia, where it is used for medical purposes. Asia’s demand cannot be met through legal channels as the Convention on International Trade in Endangered Species (CITES) of Wild Fauna and Flora[4] banned the international trade in rhino horn in 1977, when it became apparent that the high levels of poaching would ultimately lead to the extinction of the species.[5] Challenges South Africa has undoubtedly made great strides in curbing poaching and protecting its rhino population. There are numerous public awareness campaigns focused on the cruel slaughter of these animals and the media often covers reports of arrested/convicted rhino poachers. However, no durable solution has yet been found.. Although, poachers are heavily fined and, in some instances, jailed, this happens only after the killing of another rhinoceros. There are several challenges facing anti-poaching initiatives and in particular the development of devices that can monitor the behaviour of rhinos. For one, the vastness of their habitat hinders attempts to prevent rhino poaching.[6] Further, the anatomical structure of the rhinoceros renders ex vivo attached devices to monitor and study them difficult. The difficulty of establishing wireless communication with these devices and the impact of territorial combat on animal-attached devices further exacerbate the problem. Finally, “little is known about the physiological, social and migratory behaviour of the rhinoceros.”[7] This impedes the development and optimisation of any monitoring system. To address these challenges, and ultimately to facilitate nature conservation and anti-poaching initiatives, researchers at the Department of Electrical and Electronic Engineering, University of Stellenbosch are developing an animal-borne behaviour classification system. Preliminary tests have already been deployed, and the system tested on rhinoceros by attaching it to the hind legal of the animal. The system’s biotelemetry tags can, in real-time, distinguish between classes of animal behaviour (such as standing, walking, grazing, running, and lying down) using statistical classification of tri-axial accelerometer measurements.[8] A prototype of the system was able to transmit live behavioural updates approximately every 6.5 seconds. The aim is to use these animal-borne classifiers in rhino conservation applications. By monitoring rhino behaviour, the aim is to automatically detect a sudden, unexpected change in behaviour that may signal an imminent poaching attack, thereby providing an early warning that may save the animal. A key challenge is the provision of sufficient power to keep the tracking system active long-term. It is not viable to tranquilise the rhino every time a battery change is required as this poses a significant risk to the animal and the staff as well as being expensive. To address this challenge, research is being conducted to develop energy-aware feature and model selection techniques for the onboard behaviour classification.[9]  Furthermore, using kinetic energy harvesting for energy self-sufficiency is being explored. The harsh reality, however, remains: There is no single magic solution that can solve the poaching crisis. It will take the best tools we can design and continued funding to keep this beautiful species from extinction. To find out more about this project contact Thomas Niesler trn@sun.ac.za or Riaan Wolhuter  wolhuter@sun.ac.za [1] https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0216595#sec001 [2] https://www.savetherhino.org/rhino-info/poaching-stats/ [3] https://www.savetherhino.org/rhino-info/poaching-stats/ [4] Adopted on 3 March 1973, entered into force 1 July 1975. 993 UNTS 243. [5] https://www.environment.gov.za/sites/default/files/docs/rhinohorntrade_southafrica_legalisingreport.pdf [6] https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0216595#sec001 [7] https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0216595#sec001 [8] http://dsp.sun.ac.za/~trn/reports/leroux+marais+wolhuter+niesler_animalbio17.pdf [9] https://www.researchgate.net/publication/329661763_Energy-Aware_Feature_and_Model_Selection_for_Onboard_Behavior_Classification_in_Low-Power_Animal_Borne_Sensor_Applications.

Don’t miss the Engineering Open Day 2019 on 23 Februarie 2019 @ 08:30 – 16:00

Well done to our local celebrity, Dr. Herman Kamper, who discussed some of the main aspects about the group maties, Maties Machine Learning, our local Machine Learning and Artificial Intelligence (AI) interest group. This group will also be hosting the biggest AI conference in Africa in 2018 here in Stellenbosch the Deep Learning Indaba. Also learn more about the group from this SU post Watch the video here:

The 2018 edition of the biennial conference of the South African IEEE, Joint Chapter on Antennas and Propagation (AP), Microwave Theory and Techniques (MTT) and Electromagnetic Compatibility (EMC), will be held in Stellenbosch on August 30 and 31, at the Stias Conference Centre. This conference brings together engineers and researchers from industry and academia who work in fields related to AP, MTT, and EMC in South Africa. All conference contributions will be by invited presenters, including a few international, plenary speakers. The conference is organized by the South African IEEE Joint AP/MTT/EMC Chapter. See the conference website for further details.

On 15 November, 2017 the final year EE students showcased their final year Skripsie projects. The innovation that was illustrated by these young Engineers was absolutely breathtaking. Here are some pictures of the day. We would also like to congratulate our representative, Ryan Eloff that represented us on 6 December at the annual Jac van der Merwe Competition for Innovation, sponsored by MultiChoice. His final year project was entitled, Teaching a robot to interpret natural language navigation instructions.

As from 2018, the Department of E & E Eng is offering a new Structured M.Eng programme in Smart Grid Technology. This is a new course offering in response to worldwide evolutional processes in the electrical energy domain. These are seen to be technologically very exciting, but will also have considerable impact on conventional networks, in the near to medium future. Administrative / Academic Requirements: Prerequisite: Any four year engineering B-degree, any BTech degree in engineering, or a BSc degree with a major in Physics Duration and Teaching Load: One, or two years; Full time- or part-time basis.  One week block per module with 45 hours of contact and additional work via distance education. Successful completion of all modules is followed by a thesis project, as set out elsewhere. Each block carries 15 credits and the project 60 credits. Modules are presented on both NQF levels 8 and 9, but at least four modules must be taken at NQF level 9. Course Content: Details of the content in the different course modules, are available on this page. How to Apply: The application procedure for this programme, can be found on this page

CapeTalk and Shoprite have partnered in a special campaign aimed at saving water at schools across Cape Town. The campaign involves the installation of a special water monitoring device called the Dropula, at schools. This device is designed by Prof Thinus Booysen, a lecturer at Stellenbosch University. The Dropula monitors water usage and flow. It is a smart meter that attaches to the municipal meters that supply the school. The information is disseminated and sent to the teachers, head masters and the children for them to see their consumption level. The school will also have access to a dashboard which alerts them to a sudden changes or problem with water usage or flow. Once there’s an alert a maintenance team is dispatched to the site (school, business etc) to pinpoint the fault. The device has been installed in 6 schools to date. You can read more about the water saving initiative for the schools at www.schoolswater.co.za. Have a look at the interview on Cape Talk. You can also read more about Prof. Booysen’s company behind the Dropula at www.bridgiot.co.za.