John is an R&D Engineer at MMC-Gamuda KVMRT (T) Sdn Bhd. He is currently working on the Klang Valley Mass Rapid Transit (KVMRT) Sungai Buloh-Serdang-Putrajaya Line 2 megaproject in the heart of Kuala Lumpur. He leads the research effort into autonomous TBMs together with a fantastic team.
John graduated from the University of Bristol with an MEng in Mechanical Engineering with a strong research focus on machine learning. He is fluent in MATLAB and Python with an intermediate knowledge in C. Through the years, he has worked on various projects including developing AI algorithms for a chess engine, localisation and path planning algorithms for robots, applied machine learning to cancer microscopes and manifold-learning algorithms to drone landings.
Topic: Using Python to develop Autonomous Tunnel Boring Machine Algorithms.
Tunnel Boring Machines are large moles in the ground which excavate underneath our busy streets without any disruption to the surface. A TBM is equipped with over 400 sensors which are linked to a Siemens PLC for process control. Traditionally, an operator is required to sit in a control cabin within the machine to steer and control various parameters to ensure a smooth and safe mining process. With so many sensors, a human operator is data overloaded and has to look at each data point sequentially with little room to optimise.
Our Autonomous TBM innovation is set to revolutionise the industry by fully controlling a machine using control algorithms written in Python. Computers are adept at the task of data processing and making calculated decisions. Using the power and flexibility of Python, we have been able to quickly innovate and develop this system with the existing tools and libraries available. This innovation was developed on the local Klang Valley Mass Rapid Transit Project SSP Line 2. We have successfully tested this innovation on two machines from Bandar Malaysia to Chan Sow Lin. This talk will cover the project, the innovation and a short tutorial on PLC communications with Python.
Targeting audience with intermediate level of Python knowledge.