We are hiring across software, hardware, and business operations roles for robotics applied to grouting and the Fabricator reinforcement learning platform.
Interns may apply for semester-long internship (credit-bearing), full-time for at least 3 months, or in very exceptional cases, part-time for at least 5 months. Generally, University Interns get S$1500 a month in gross salary or S$1755 without CPF for full time work, prorated for part time OR portion in share options. We are open to polytechnic and high school students as well.
If you qualify for our internship under the SGInnovate summation program, you may receive a stipend of S$3000-S$6000 (checkout Fabrica AI's projects on their website for more information)
Join us in shaping the future of robotics
Why us
Open positions in Czechia (hybrid/remote work possible)
Requirements
-
1 year of any experience with deep learning in Python
-
Interest in reinforcement learning applied to robotics
-
3 years of experience in C++ or Python development
Nice to have
-
Any further general software development experience in C++ or Python
-
Robotics experience
Who are you?
-
A brilliant engineer that likes Python code that spends the minimum amount of time in the interpreter
-
You pride yourself in seeing physical impact, not only higher ads revenue
-
Some experience with deep learning (eg. CV, NLP, general classifiers, entity extraction, etc.)
-
Experience in version control (git - branches, pull requests workflow)
-
Willing to work with Agile development methodology
-
Experience with Continuous Development / Continuous Integration (Github pipelines)
What will you do?
-
Developing logic and simulations in Python and Pytorch for Isaac Sim
-
Solve problems and eliminate bottlenecks – GPU and distributed computing
-
Write Isaac Sim plugins for other developers using the Fabricator
-
Integrate pre-trained transformer models, photo scans, or CAD programs
-
Make the project development more effective
-
Help in the deployment to the grouting robots
-
Requirements
-
2 years of applied experience with Deep learning in Python
-
Worked on a deep learning focused project
-
-
Prior experience with reinforcement learning (work, thesis, personal project)
-
Interest in reinforcement learning applied to robotics
-
Experience in version control (git - branches, pull requests workflow)
-
Willing to work with Agile development methodology
Nice to have
-
Any general software development experience in C++ or Python
-
Robotics experience
-
Experience with Continuous Development / Continuous Integration (Github pipelines)
Who are you?
-
A brilliant engineer that likes Python code that spends the minimum amount of time in the interpreter
-
You pride yourself in seeing physical impact, not only higher ads revenue
-
Hands on experience with deep learning (eg. CV, NLP, general classifiers, entity extraction, etc.
What will you do?
-
Developing RL and logic in Python and Pytorch for Isaac Sim
-
Solve problems and eliminate bottlenecks – GPU and distributed computing
-
Delve into rewards, metrics and environments
-
Improve convergence on all levels
-
Turn CAD models into conceptual designs
-
Requirements
-
Previous research experience in Deep learning or reinforcement learning (academic or non academic)
-
Knowledge of the Deep RL landscape
-
2 years of experience with Deep learning in Python
Nice to have-
Robotics experience
-
Software engineering experience
Who are you?-
A brilliant autonomous researcher in the field of Deep RL
-
You pride yourself in seeing physical impact, not only higher ads revenue
What will you do?-
Developing RL and logic in Python and Pytorch for Isaac Sim
-
Conduct research on state-of-the-art reinforcement learning algorithms and techniques in robotics
-
Develop novel RL methods for custom hardware design and control optimization
-
Collaborate with the engineering team to translate research findings into practical solutions
-
Publish research findings in top-tier conferences and journals, showcasing the company's expertise in the field
-
Stay current with the latest research in reinforcement learning, machine learning, and robotics to drive innovation within the company
-
-
Work on applied deep reinforcement learning project
-
Tackle open ended hardware design search
-
Aim to push the state of the art
-
Test the RL policies on our physical grouting robots, where we have access to all hardware and software
-
Results will be deployed on Singaporean construction site, so you gain applied RL experience
-
Have a lot of freedom to influence the course of the development
-
Small, learning focused Fabricator team with resources for professional growth
-
The option to get significant stake in the company
Many tasks are not automated, because of the cost and uncertainty in building robotic solutions.
We aim to move the needle on building robotic solutions by automating the hardware design itself. We are building a platform called Fabricator, which aims to make the design process 100 times faster. The idea is to create robots automatically by defining their tasks and operating environments. Right now, we're working on the first stage with the Early Fabricator. It's a simulation tool that helps decide the basics of robot design like shape, mechanics, sensor placement, and drivetrain selection. This could cut down the early design phase for robotics companies by half a year to a full year.
Example of using Fabricator to develop the design of a Grouting robot:
Fabricator TL;DR
If you come on board, you'd be working on the Early Fabricator, which is currently under development with the aim to roll out to other companies by mid-2025. It's a real chance to dig into applied deep reinforcement learning, push the boundaries of hardware design, and see your work in action on real robots that are getting their wheels dirty on construction sites.
We understand this is a hard and research oriented project. To stay grounded in reality, we are our own reference customer. In under 2 years, we've developed a working, patented and profitable grouting robot whose version 4 was optimized by the proof-of-concept of the Fabricator in late 2022.
Hear from our employees
Do you want to know what it is like working at a such a dynamic and innovative tech start-up? Hear it from our employees and why they love working for Fabrica Ai!
Your internship would revolve around real-world industry deployments taking us from a TRL of 7 to 9. You may be remunerated with a combination of salary and share options. The following tracks are available, you would probably be doing a couple of them up to your preference, with examples of what has been done by previous interns (plus most of assembly has been done by interns):
1)Mechanical experiments and design
-
Experiments - e.g. designing variations to the grout extruder to allow less waste and better reliability
-
Design - e.g. most of the upcoming v4 robot is being designed by interns with Jakub supervising them
2) Mechanical manufacturing, assembly and testing
-
3D printing - Slicing prints (often for multi-filament printing), managing and repairing printers (we have Tenlog TL-D3, TL-D5, Ender 3, Prusa i3)
-
Outsourcing manufacturing - specifying cabling, machining etc. to manufacturers
-
Machine tools - making some parts by hand when we need to quickly iterate
-
Assembly - putting in inserts, bolting parts together, waterproofing
-
Testing - testing the robot’s grout extrusion and cleaning, making minor adjustments to make sure everything is running properly and robot produces nice grout lines
3)Electrical experiments and design
-
Design - most of our PCBs have been designed by interns, aiming to minimise free wires in the robot, speed up assembly, allow easy modification where needed. Generally connecting all modules, controlling power, providing a physical user interface…
-
Experiments - e.g. sourcing and testing various motor and driver combinations to pick for our drive system
4) Electrical assembly
-
Outsourcing production of cable assemblies and PCBs
-
Where necessary to make modifications - soldering PCBs, soldering power cables, crimping cables
-
Embedded programming
-
Writing object oriented C++ code
-
Keeping the code clean and maintainable
-
Your internship would revolve around bringing our robot to market. You may be remunerated with a combination of salary and share options. It is possible to combine this with our technical internships for a more “hybrid” role. This internship requires a reasonable amount of technical understanding/engineering interest.
-
Assisting with the recruitment process, including posting job openings, reviewing resumes, and conducting initial candidate screenings.
-
Coordinating and scheduling interviews with job applicants and hiring managers.
-
Assisting in onboarding new employees, including paperwork and orientation.
-
Maintaining and updating employee records and databases.
-
Assisting with HR-related communications, such as preparing HR documents and announcements.
-
Supporting HR team with various administrative tasks.
-
Assisting in the development and implementation of HR policies and procedures.
-
Conducting research on HR trends and best practices.
-
Scheduling and joining meetings with prospective clients, investors, and partners
-
Accounting and purchasing
-
Scouting for new opportunities, writing proposals for them, and pushing fabrica out there
-
Visiting industry trial sites, giving appropriate feedback to product development team
-
Speaking to industry players, ground workers, understanding the process of tile grouting to help us shape our solution, especially from an operational point of view
-
Connecting manufacturers and our product team to build and iterate on our prototype, creating new avenues for manufacturing at scale, manufacturing for reliability
-
Working with Manufacturers for Design for Manufacturing, reliability, and assembly
-
Hiring and recruitment
-
Website and other marketing/public relations management
-
Any other operational duties
-
Your internship would revolve around real-world industry deployments taking us from a TRL of 7 to 9. You may be remunerated with a combination of salary and share options. The following tracks are available, you would probably be doing a couple of them up to your preference, with examples of what has been done by previous interns (plus most of assembly has been done by interns):
1) (Software) Create new data generator for object detection (ideal is some experience with Blender/Unreal/Unity)
2) (Deep Learning and Machine Learning): Train, and evaluate our computer vision model fine-tuned on synthetic data on a real-world test set.
3) (Deep Learning engineering): Go through the list of researched libraries, use them to integrate the model into ROS2, and evaluate (compute) performance. Test performance uplift on low-power DL co-processors.
4) (algorithms) Write an algorithm to turn object detection bounding boxes in pixel space to 3D space using known constraints (baseline algo already researched). Turn it into a ROS2 node.
5) Robotics DevOps & Real World Industry Trials (must be in Singapore)
-
Understand the whole project and run our ROS2 stack on real HW.
-
Create reproducible testing methodology and evaluate metrics on the HW.
-
Be in charge of the robots on the construction site.
-
Explore and procure new parts/new materials for robot functionality improvement
-
Core responsibilities:
-
PCB, electronics design
-
Working with both hobbyist and industrial microcontrollers, single board computers, motor drivers etc.
-
Integrating motor systems of different types (DC, BLDC, with encoder feedback…), sensor modules etc.
-
PCB design
-
-
Embedded programming
Possible additional responsibilities:
-
Testing our robot on construction sites
-
Managing external PCB production
-
Harness design
-
Overseeing a team doing all the above
-
-
Electrical experiments and design
-
Experiments - e.g. sourcing and testing various motor and driver combinations for our main drive system
-
Design - e.g. designing PCBs to replace a mess of wires and extend the variability and power of the control electronics; PCBs for protecting the AC power supply from back-emf voltage spikes
-
-
Electrical assembly
-
Soldering PCBs and power cables, crimping cables
-
-
Arduino programming
-
E.g. Writing code for TMC2209 to reduce power usage of our stepper motors and replace force sensors with smart code
-
Controlling and calibrating the extrusion and cleaning systems
-
-
Robotics DevOps & Real World Industry Trials
-
Understand the whole project and run our ROS2 stack on real HW
-
Create reproducible testing methodology and evaluate metrics on the HW
-
Be in charge of the robots on the construction site.
-
Explore and procure new parts/new materials for robot functionality improvement
-
-
Open positions in Singapore
Open positions in Czechia (hybrid/remote work possible)
-
Work on ROS2 foxy nodes logic/topics for object detection with the main camera, Floor segmentation, Feedback-loop based navigation correction
-
Auto calibration of cameras, LIDAR (position, noise, hardware characteristics)
-
Advanced procedural Gazebo simulations to automatically evaluate the robots performance on suite of metrics
-
2D box physics based simulations for Reinforcement Learning (RL), setup and run RL experiments
-
Our ideal candidate
A great candidate would possess the following:
-
Willing to work without a very well defined job scope and willing to adapt to changes week on week
-
Strong algorithmic thinking
-
Ability to think of metrics to measure progress
-
Understanding of computer science fundamentals, schema design
-
Eagerness to stick to best practices
-
Value the direct impact on the product and proximity to clients at a startup and the growth potential of a startup
-
Growth mindset driven
-
Motivation to build a great product and codebase using the most recent tech stack for the slowly coming robotics revolution
-
Deep learning applied to more than ad revenue, end2end reinforcement learning, etc.
-
Insert other cliche adjectives from other job posts
Our development process is very iterative and dynamic. Thus, there would not be a concrete plan more than 2 weeks ahead and your role/tasks is/are likely to evolve over time. If you like this process, you’d be a good fit for us. You would thus play a significant part in deciding how your work would contribute to our overall goals. There are some exceptions to this such as our Gazebo simulation, which has already been verified.
If you are applying for an internship, we are looking for a minimum duration of either 3-months full-time or, in very exceptional circumstances, 5-months part time. Internships are open all year-round with rolling closes, generally closing a few months in advance. For example, our summer internships for 2023 are filled around March.
Please email your Resume to careers@fabrica.ai with an introduction of yourself, the role(s) you are applying for, and the time period you are applying for if you are interested. Thank you.
*If you are interested in the software positions in our Czech office but are based elsewhere, you may still send in your application.
Grouting robot
We invented an automatic tile grouting robot that boosts productivity by up to 5x and are piloting it on construction sites in Singapore managed by Woh Hup, BHCC, Gin Chia, Soilbuild, Chan Rong Fen, and Dragages. We have completed over 100,000sqft of space to date (late 2024) and have started commercial deployment and international demos. Tile grouting is the process of filling up the gaps between tiles after the tiles have been laid. We invented a patented robot that is able to automatically grout and clean tile gaps with only a press of a button.