About
Founder & CEO of Matta - a startup creating industrial AI for factory…
Activity
4K followers
Experience
Education
-
University of Cambridge
-
-
- Thesis title: "Deep learning enabled error detection and correction for 3D printing"
- IET Postgraduate Scholarship for an Outstanding Researcher
- Best presentation at IfM First Year PhD Conference
- IET National Travel Award -
-
-
-
-
-
Volunteer Experience
-
Mentor
Student Robotics
- 5 months
Education
Student Robotics is an organisation which runs robotics competitions for 16-18 year olds. I supported the school children in designing their robots, aiding them with programming and electronics. Also I assisted with the planning of the competition and tasks in frequent volunteer meetings.
Projects
-
OctoRest
- Present
See projectOctoPrint is a popular web interface for 3D printers allowing users to control and monitor their printer through the browser. OctoRest is a Python wrapper for OctoPrint's REST API. Simply install the package via PyPI or clone the repo to add communication with OctoPrint to a project.
-
3D Printed Crystal Palace
-
See projectI designed and built an absolutely massive fully 3D printed model of The Crystal Palace from the Great Exhibition of 1851. All the designs were copied from the original architectural drawings which I examined in the V&A and Royal Commission for the Exhibition of 1851 archives. 5 3D printers were then printing non-stop for about 10 weeks to produce all of the parts which were then assembled. Over 40,000 people viewed the model on the exhibition weekend in South Kensington, London.
-
FFF 3D Printer Error Detection (Thesis)
-
See projectFused Filament Fabrication (FFF) is the most popular low-cost 3D printing technology. However, currently all consumer FFF printers have limitations in reliability as they operate in an open-loop fashion. Both time and materials are wasted due to the frequency of failed prints and the machine being oblivious to errors occurring. This work focuses on addressing this issue by adding computer vision based in situ process monitoring to FFF printers. The proposed methodology can be used with any…
Fused Filament Fabrication (FFF) is the most popular low-cost 3D printing technology. However, currently all consumer FFF printers have limitations in reliability as they operate in an open-loop fashion. Both time and materials are wasted due to the frequency of failed prints and the machine being oblivious to errors occurring. This work focuses on addressing this issue by adding computer vision based in situ process monitoring to FFF printers. The proposed methodology can be used with any existing machine and utilises a low-cost camera to detect printing failures in their infancy. The system employs both passive and active approaches to monitor the print. The former method purely observes and analyses the object unaware of its desired shape, whereas the latter directly compares the monitored print to the expected 3D model. When an error is detected a user is notified and/or the print terminated. Rapid failure detection leads to reduced time spent on erroneous prints and less waste material. Results show high accuracy in detecting the most common and significant errors (93.8%) and correctly classifying successful prints (80.6%).
-
VisUAL2 - ARM Assembler and Simulator
-
See projectVisUAL2 is a cross-platform tool that makes it very easy for beginners to write small and medium-sized ARM assembler programs, with no manual and intuitive visual debugging. Even those very familiar with ARM assembler have found VisUAL2 useful for quick writing and debugging. The IDE has helpful error messages and an ergononic GUI that makes debugging visually intuitive. VisUAL2 is a re-engineered and improved version of the original VisUAL created in 2014. Both VisUAL2 and VisUAL are written…
VisUAL2 is a cross-platform tool that makes it very easy for beginners to write small and medium-sized ARM assembler programs, with no manual and intuitive visual debugging. Even those very familiar with ARM assembler have found VisUAL2 useful for quick writing and debugging. The IDE has helpful error messages and an ergononic GUI that makes debugging visually intuitive. VisUAL2 is a re-engineered and improved version of the original VisUAL created in 2014. Both VisUAL2 and VisUAL are written by Imperial College students.
-
Full-sized humanoid robot
-
See projectAnakin Skywalker’s creation of C-3PO as a boy awakened the challenger within me. Yes, he lived in a galaxy far, far away but could I build one too? Inmoov, is an open source 3d printed robot with an excellent community. Over the past couple of summers, I have been constructing this android, called Flanders, and contributing in a small way to the community by introducing improvements to the build. After much 3d printing with my self-constructed RepRaps and assembly, I am currently implementing…
Anakin Skywalker’s creation of C-3PO as a boy awakened the challenger within me. Yes, he lived in a galaxy far, far away but could I build one too? Inmoov, is an open source 3d printed robot with an excellent community. Over the past couple of summers, I have been constructing this android, called Flanders, and contributing in a small way to the community by introducing improvements to the build. After much 3d printing with my self-constructed RepRaps and assembly, I am currently implementing functionality by writing image processing and inverse kinematic applications so Flanders can interact with objects.
The servo motors for many of the joints had to be modified to avoid poor code and to enable continuous rotation. The potentiometers on some had to be reversed, whilst other movements needing two servos, such as the waist, required both servos to use the same potentiometer.
Six servos are housed in the forearm. The hand is actuated like human hands using tendons, which control the fingers and wrist. These were made from high-tensile fishing wire on both sides of a finger. As a servo turns it loosens one tendon, whilst pulling the other, resulting in movement.
Another feature I have implemented is the tracking of human faces, so that the robot can move its head and look at you. This gives the robot a far more life like appearance, although challenges were encountered when presented with multiple faces.
An uncle of mine lost his hand as a child and I pondered, how hard would it be to create a usable prosthetic hand? The creation of a hand itself was not too challenging, the main task was how to interface it with the human and the required control mechanism. I achieved a useful level of functionality with grasping, opening and moving left and right, all controlled using the electrical impulses in the users arm. -
FPGA Based Augmented Reality (AR) Headset
-
See projectDuring first year at university, in a team of three we got very interested in the field of FPGAs and their possible uses in augmented reality. Please see our final report on our work, poster and a video about it. What is that hair Douglas!!! We built the hardware and wrote the software for the full solution and won best 1st year project.
With virtual and augmented reality markets (VR/AR) expecting to reach a value of £130 billion by 2020, three of us were fervent to innovate in this…During first year at university, in a team of three we got very interested in the field of FPGAs and their possible uses in augmented reality. Please see our final report on our work, poster and a video about it. What is that hair Douglas!!! We built the hardware and wrote the software for the full solution and won best 1st year project.
With virtual and augmented reality markets (VR/AR) expecting to reach a value of £130 billion by 2020, three of us were fervent to innovate in this space. Another growing technology, FPGAs, are integrated circuits designed to be configured by the consumer. These allowed us to develop a novel solution for tracking markers, applying matrix transforms and achieving high refresh rates for AR systems in what we believe was one of the first demonstrations of FPGAs for such an application. We designed the headset, PCB, drivers, HLS tool and applications for the headset and received funding from the Xilinx University Program. Truly a full stack engineering project! We achieved frame rates exceeding 130fps. The two cameras on the front were used to track markers in the real world. Inverse matrix transforms were used to project content against a surface. The content was then overlayed on the video stream from the cameras and displayed on the 4.8inch LCD.
The project’s ability to stably detect markers and render a live transposed image over them gives the user a platform to allow creative applications to be developed. To improve accessibility we developed packages for text editors to assist programming in Elastic C, our own language developed for the project.
We developed three applications during the project; 2D Minecraft with mining and placing of blocks, HDMI overlay on a surface, and a live drone camera feed on surfaces.
After completing the project we produced the following poster to summarise what we achieved to inspire prospective students to come and study at Imperial.
Honors & Awards
-
IET Postgraduate Scholarship for an outstanding researcher
IET
-
IET National Travel Award
The Institution of Engineering and Technology (IET)
In early April 2022 I was awarded the IET National Travel Award to attend the “Special Event: BMVA 3 Day Symposium” in Manchester. This event was run by The British Machine Vision Association and Society for Pattern Recognition to bring together all machine vision researchers based in the UK through a comprehensive programme of keynotes, "missed" orals from recent online conferences, and extended poster sessions. Due to the global pandemic, during the first 2 years of my PhD at the University…
In early April 2022 I was awarded the IET National Travel Award to attend the “Special Event: BMVA 3 Day Symposium” in Manchester. This event was run by The British Machine Vision Association and Society for Pattern Recognition to bring together all machine vision researchers based in the UK through a comprehensive programme of keynotes, "missed" orals from recent online conferences, and extended poster sessions. Due to the global pandemic, during the first 2 years of my PhD at the University of Cambridge, I have been unable to visit any in person events or conferences. Therefore this symposium was a fantastic opportunity to present my work in person and learn about the latest research being undertaken in the UK. Most importantly, this event provided an opportunity to meet the community and establish relationships for future collaborations – whilst also having a good time with other machine learning researchers.
Read more about my trip:
https://www.theiet.org/impact-society/awards-scholarships/travel-awards/travel-awards-for-national-travel/winners/ -
Imperial Venture Catalyst Challenge AI & Robotics Winner
Imperial College Enterprise Lab
Winner of the AI & Robotics track part of Imperial’s flagship entrepreneurship competition taking home £10k in prize money. The result of an intensive 2 months of masterclasses, 1-1 coaching, and meetings with experts, which finished with a large pitch streamed live watched by 1,000s of people. In the top 5 startups from over 200 applying.
-
Runner-up in Cambridge Entrepreneurs software startup competition
Cambridge University Entrepreneurs
-
Best presentation at IfM First Year PhD Conference
-
-
Royal Commission for the Exhibition of 1851 Industrial Design Scholarship
Royal Commission for the Exhibition of 1851
For outstanding engineering or science graduates who wish to develop their capabilities in industrial design and who aspire to becoming leading designers in British industry. The scheme is open to applicants resident in the UK who are intending to make a career in British Industry.
-
Dean’s List
-
For graduating in the top 10% of the year.
-
Governors’ Prize
-
For the outstanding BEng student in Electronic & Information Engineering.
-
Runner-up and finalist in Durham Blueprint competition
-
-
Best 1st Year Project Prize
-
For the best 1st year FPGA image processing project. Created an FPGA based AR headset with custom PCB with HDMI in. Developed Elastic C, a specialised HLS tool for low latency streaming applications, going from a C like language to a single cycle, pipelined implementation. 3d printed AR headset and developed a version of Minecraft to ‘snap’ to surfaces by applying a perspective transform matrix. Received donations from Xilinx and worked in a lab at Imperial on 4k AR.
-
Ash Music Scholar
-
1 of 5 music scholarships offered per year between Imperial College London and the Royal College of Music.
-
Rotarian Young Achiever Award
The Rotary Club
In recognition for many outstanding achievements.
-
Academic Scholarship
Hurstpierpoint College
-
Design & Technology Scholarship
Hurstpierpoint College
-
Music Scholarship
Hurstpierpoint College
Other similar profiles
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More