Poster Session

Presented by Zaynab Awofeso and Yusuf Olaniyi

Damages from cracks in reinforced concrete (RC) structures have led to the loss of lives and property globally. This has led to numerous studies in the area of structural health monitoring (SHM). SHM processes are typically divided into four stages: damage detection, location, extent, and prognosis of structural damage. Given the labor-intensive and time-consuming nature of traditional visual inspections, our research aims to harness the power of artificial intelligence, specifically advanced convolutional neural network (CNN) models - YOLOv5, FastRCNN, and Single Shot Detector for crack detection in RC structures. This initiative is being conducted using the structural facilities at the University of Lagos for a start.

The notable gap also in the availability of local datasets for crack detection prompted the initiative to collate and annotate a unique dataset of 1000 instances of cracks specific to Nigerian infrastructure. To date, 322 images of cracks have been collected and annotated using Roboflow to contribute to building a robust dataset for the research community worldwide.

Preliminary phases of the research have involved data preprocessing and the commencement of the initial implementation of the proposed CNN models. Future work will evaluate these models based on precision, recall, and overall accuracy metrics. Upon completion, the most effective algorithm will be deployed via an IoT-enabled sensor system, making it applicable for real-time, automated visual inspections of structural integrity to ensure safety and durability in RC constructions.

Presented by Florian Bednarski and Alesha Lamont

On our poster, we will describe how researchers at the Early Learning Lab (University of Auckland) use Pygame, to design interactive eye tracking experiments for infant participants. Specifically, we have developed a dynamic game where infants can reveal hidden images by simply moving their gaze across the screen. The game is structured by a Pygame event loop and utilizes input from an eyetracker. Developing experiments with Python offers researchers unprecedented control over the experimental process, while offering accurate data that is automatically organised and stored. This study will provide novel insights into the way that infants explore their visual field and learn from the world around them. This unique project would not be possible without the adaptable tools that Pygame offers.

By harnessing the flexibility and accessibility of Python, our team has overcome the limitations often imposed by proprietary software. In our poster presentation, we will outline how Python empowers researchers in the scientific community to create unique study designs while adapting them to evolving research needs without constraints. Furthermore, our approach prioritises transparency, as Python enables us to work with raw data measurements directly, eliminating any obscured processing or filtering.

In addition to its technical advantages, Python's open-access nature fosters a culture of collaboration and reproducibility within the research community. Our poster presentation will highlight the ease with which Python-based experiments can be replicated and shared. Utilizing version control platforms can significantly advance knowledge distribution and help to achieve scientific progress. We hope that sharing these unique uses of Pygame in a scientific context will inspire more researchers to use the valuable Python tools available to create new innovative research procedures in their own fields.

Presented By Alain de Verneil

Over the past 20 years ocean gliders have become a popular tool for collecting oceanographic observations. Since the end-users are primarily a community of scientists who are not traditionally trained in computer science or coding, the freely available post-processing code is sourced from individual research groups whose scope reflects their individual use case. Here at NIWA, we have identified the need to bridge gaps between existing packages to fulfill our desired specifications. Therefore, we're presenting KiwiGlider, a python package for New Zealand's gliders that incorporates these codes and capabilities converted from other computing languages to provide end-to-end processing from raw data to ready-to-upload database-compliant outputs. A real dataset is used to present an example data processing pipeline.

Presented by Ben Denham and Grant Paton-Simpson

Lean Python is a community initiative to shrink Python so it better fits our brains. The strategy is to provide guidance on when we should use particular language features (and when we should not). Specifically, we want to reach agreement as a community on which language features are Common Python, which are Situational Python, and which should be Deprecated.

Categories

------------

Common Python is features we should all know so we can read each other's work.

Situational Python is features that are useful but not for everyone or not all the time. Depends on situation. For example, web development needs different features than scientific Python. Advanced library code might need advanced features. For each feature provide guidance on situations where it is the preferred One Obvious Way. Situations might be defined by domain (e.g. web application development vs data science) or type of code (e.g. scripting vs code with enterprise responsibilities).

Deprecated Python refers to features we should avoid using, and possibly remove from existing code.

Practical Steps

----------------

* Step 0 was creating the Lean Python Manifesto (known as The When Of Python).

* Step 1 was making a PoC Catalogue.

* Step 2 was to blog on numerous Python language features.

* Step 3 is to get a Lean Python PEP accepted.

Presented By Maxim Danilov
A standard Django project involves working with multiple files and folders from the start. Let's see how the work with a Django project changes itself when we have only one file in project. This solution automatically transforms Django into a microservice-oriented async framework with "batteries included” philosophy.

Maxim is unfortunately not able to attend in person but will be happy to answer questions via the #kiwi-pycon channel in Python New Zealand's Slack workspace.

Presented By De-Graft Nana Baisie Affail

We will discuss methods for programming quantum computers in our discussions.

They are programmed using the Qiskit Python framework, which is used to construct quantum circuits.

We shall define the quantum communication protocol QBC (Quantum Bit Commitment) and the quantum problem USD (Unambiguous State Discrimination) in order to illustrate a few techniques.

In order to carry out the QBC experiment, we plan to create a class that abstracts this protocol.

We shall demonstrate its functioning using the QBC's BB84 and BC90 implementations

Likewise, regarding the USD, we will briefly discuss how memory-effective POVM (Positive Operator-Valued Measure) measures are implemented on quantum computers

Presented by Ben Denham and Grant Paton-Simpson

At KiwiPycon 2023, a plan was hatched to help professionals learn Python. One year later, over sixty learners have dived into Python at PyNoon lunchtime training sessions run from open-source training materials made available at https://pynoon.github.io

PyNoon is a free community training initiative that has launched the wider technoon.org movement that aims to help professionals learn technical skills with individual support in a friendly learning environment.

This poster will focus on:

1. Presenting the PyNoon vision and what it uniquely offers learners within the wider context of online courses, corporate training, and tertiary education.

2. Reporting on the success of PyNoon training in its first year, including an inaugural 10-week course attended by learners from 10 organisations and a second company-internal 4-week data-focussed course.

3. A call to action for learners to attend a PyNoon course and for Pythonistas to host a PyNoon themselves using the ready-to-go training material.

Presented by Shivay Lamba and Gaurav Pandey

This poster will cover how WebAssembly is moving beyond the browser and is pitched to become a foundational element of cloud-native server apps and serverless applications.

Python and WebAssembly pair extremely well as WebAssembly helps with the extensibility of Python. In recent years we have seen an explosion of usage of Python in the browser with projects like Pyodide, PyScript, etc. All of this is possible thanks to the powerful functionalities of the underlying platform, WebAssembly. Similarly Webassembly has also seen a huge rise in serverless and server side due to it's small size in comparison to containers, thus it can make edge deployments for Python much more efficient and quicker for serverless workloads due to it's low cold start time.

Let's explore how Webassembly benefits Python by looking how it is being used to power Python in the browser using projects like PyScript and in the serverless space with projects like Spin.

Shivay and Gaurav are unfortunately not able to attend in person but will be happy to answer questions via the #kiwi-pycon channel in Python New Zealand's Slack workspace.