The Research & Development is part of an ongoing collaborative partnership between Waldeck, Nottingham Trent University, Birmingham City University and Network Rail. Muhammad is part of the team researching and developing our joint in-house asset viewing and examination solution for industry.
In collaboration with Nottingham Trent University, the Knowledge Transfer Partnership (KTP) facilitates the development of the IDSES, an Intelligent Digital Structure Examination System, that integrates machine learning, deep learning, and state of the art visualisation technologies to automate and enhance examination of infrastructures, including bridges, highways, and road structures.
Muhammad is an experienced graduate having undertaken software development with companies in a variety of industries and sizes using a diverse set of technologies.
We caught up with Muhammad to find out more about his past 12 months as part of the team:
Hi Muhammad! Tell us a bit more about your career history so far?
Before joining Waldeck, I have been fortunate enough to work in a few different industries during my career so far. I have experience working with sales / inventory management software providers and the automotive industry. I also developed an interactive education mobile application for an international organisation.
What are you currently working on as part of the KTP project?
Currently, I am working on developing EVE (Examination & Visualisation Environment), which is a commercial asset examination and visualisation management software that Waldeck will provide to the wider industry. As part of this software, I have been developing both frontend and backend code for the data management part of the software as well as integrating the asset visualisation software into one whole solution.
I am also working with the team on developing the custom object detection model that can detect different types of defects and label them, saving time and improving accuracy of digital examinations. As part of this, I was responsible for labelling a large dataset for training with the different defects annotated. I also carried out research on using different object detection models to benchmark the task and carry out a feasibility study for this approach.
What will this solution provide to industry once ready?
This solution will provide the industry a way to visualise a diverse set of data through one visualisation platform: The software is able to integrate point cloud data, 360-degree imagery and flatimages to accurately provide examiners with a view of the asset so that defects and issues can be easily detected.
The AI integration will allow state of the art Computer Vision models to be available to examiners and engineers looking at assets: These will aid them in detecting defects automatically and providing detailed descriptions of the defects for the reports. This will save the end users time whilst also making the detection and labelling of defects more standardised.
Furthermore, the software will integrate state of the art Large Language Models (LLMs) to create a Knowledge Base from the previous reports and various examinations about the asset: This will allow the end users to quickly query the current state of the asset while allowing salient points from previous reports to be clearly laid out. These LLM models will also allow the generation of detailed automatic reports that give an accurate snapshot of the current state of the asset, using the Knowledge Base that is available to the model.
What do you enjoy doing outside of work?
I enjoy building things from scratch outside of work. I have deep interest in keeping up to date with Deep Learning and have built projects around this topic. I also spend a lot of time learning about developing operating systems from scratch and procedurally generating worlds and objects for computer graphics applications.
Currently, I am writing a two-volume book that teaches readers how to use basic mathematics to create entire procedurally generated scenes from scratch, the books go in depth on both the theoretical and practical aspects of this process. Readers are step by step introduced to different concepts, progressing to more complicated concepts and applications.
The book can be found here: Computer Graphics: From… by Muhammad Ahmad [PDF/iPad/Kindle]
How will your role continue to evolve at Waldeck over the coming months?
I have recently started a part-time PhD with Nottingham Trent University which will be focused on the work I do with Waldeck. A major part of my research will be on developing tools to utilise cutting edge AI technologies to aid the Infrastructure Examination process. This will allow me to utilise cutting edge research while also developing useable products and applications for Waldeck.
I will continue to work on EVE with the team over the coming months, developing it into a commercial application that can be used for multiple clients.
To find out more about our Graduates stories please click here>