About Me

I am an engineer with expertise in Artificial Intelligence and Robotics. I received my MSc degree from the University of Bristol in 2016 and I've been working in the industry since. I am interested in autonomous systems and Artificial Intelligence with emphasis on perception, decision making and control.

I enjoy being a research engineer, working on cutting edge technologies and developing systems to solve real life problems. Besides focusing on the technical challenges, I also find the ethics of technology crucial for making the right decisions as we further develop.

For an up to date employment history please see my CV or connect with me on LinkedIn.

Skills

  • Robotics
  • Artificial Intelligence
  • Machine Learning
  • Deep Learning
  • C++
  • Python
  • ROS

Projects

I contributed to a broad range of projects during my studies and work experience at the MathWorks and Cambridge Consultants. Since then I've been working as a freelancer with clients from a different background, including finance and drug discovery.

Vincent™

An AI system that turns sketches into art. It utilises a GAN architecture to mimic human creativity and imagination.

Master Thesis

Dynamic Gain Scheduling control for the Bixler aircraft in highly non-linear flight conditions. The project included simulation and control design, custom sensor design and integration and software implementation on the on-board flight controller.

Bachelor Thesis

An Artificial Neural Network based adaptive non-linear control strategy, with highlight on its capability as a compliant control scheme. The thesis involved simulation and control design, mechanical and PCB design, with custom embedded software.

Electric hokey cart

An electric wheelchair for disabled people to play indoor hokey. I worked on the control system, embedded controller and the power-electronics.

Deepray™

An AI system that reconstructs highly distorted images in real-time. Using a GAN it can go beyond simple image restoration when huge amount of information is lost. It is able to fill up missing details in a context aware way that it learnt from the training dataset.

Internship at DroneX

I worked on control algorithms and software for drones and bipedal systems. Also, I had the chance to experience working in a dynamic start-up environment.

Multidisciplinary optimisation

A satellite coverage project that involved a multi-step design process with optimisation in each step to get the best possible coverage for the given constraints.

EMICAB

EMICAB was an international collaborative research project to develop a hexapod robot. I was working on a novel tactile sensitive fingertip design for the legs.

Drone simulation

A simulation environment in Gazebo with the corresponding controller in Simulink communicating over ROS. It was developed for the IMAV 2017 student competition.

The kidnapped robot

A project to address the kidnapped robot problem, involving autonomous localisation and mapping in an uncertain environment.

Robot kinematics and dynamics

Simulation and control for several different robotic arm configurations.

Stepladder climber

A university project to build a robot that can climb a stepladder with very tight size and weight constraints.

Academic contributions

2018 Sketches into Art, V. Jankovics, Guest lecture at the University Centre Peterborough, UK.
2016 Artificial neural network based compliant control for robot arms, V. Jankovics, S. Mátéfi-Tempfli and P. Manoonpong, Int. Conf. on the Simulation of Adaptive Behavior 2016.
2014 High resolution tactile sensors for curved robotic fingertips, A. Drimus, V. Jankovics, M. Gorsic, and S. Mátéfi-Tempfli, Living Machines 2014 Proceedings.
2014 Novel high resolution tactile robotic fingertips, A. Drimus, V. Jankovics, M. Gorsic, and S. Mátéfi-Tempfli, IEEE Sensors 2014 Proceedings.