What comes to mind when you think of robots? Whether it’s an image of large-scale machinery used for industrial applications or something more domestic like an autonomous vacuum and cleaning device, there’s no doubt robots are becoming increasingly prevalent in our lives. This trend has been made possible thanks to advances in artificial intelligence and machine learning that are helping to make robots even smarter and more independent.
One Singaporean company that is focused on these advancements is dConstruct Robotics. The local start-up is a deep technology company that specializes in developing smart solutions for robots. Specifically, the company leverages AI and machine learning to capture and reconstruct 3D worlds in order for autonomous robots, drones, and vehicles to navigate safely. These solutions are especially useful in the construction industry, where robots are deployed to do mapping and surveillance at work sites, or in companies that are looking to create digital twin applications.
Bachelor of Science in Computer Science in Real-Time Interactive Simulation graduates Tng Kah Wei, Loh Xiao Qi, and Brandon Mak are part of the team at dConstruct Robotics who are putting their computer science knowledge to good use. The three DigiPen Dragons initially joined dConstruct Robotics in April 2023 as student interns before converting to full-time software engineers shortly after graduating in 2024. While all three of them specialize in graphics programming, their day-to-day work differs.
Kah Wei primarily develops features for the company’s existing robots to enhance them in support of their clients’ needs. He is currently working on two main products — dConstruct’s d.ASH Nav, which is a self-driving stack for autonomous robot navigation, and d.ASH Xplorer, a 3D mapping and reconstruction app — to facilitate data processing, map editing, and cloud interfacing. Part of Kah Wei’s work also includes speaking with clients to better understand their requests, as well as conducting on-site demonstrations to teach them how to use dConstruct’s various robots and solutions.
Brandon works on general software engineering tasks with a focus on those that are graphics related and math heavy. He frequently taps into his linear algebra and calculus knowledge for these, relying on the strong mathematical foundations he developed at school. “A lot of the things that I explored in school are very applicable,” Brandon says. “These consisted of rendering optimization algorithms, visual effects, and lighting techniques. I also have a ton of experience with OpenGL, which has really come in handy for my current job.”
In Xiao Qi’s case, he focuses on developing tools and libraries that other engineers can use. For example, he developed an app using his own custom rendering engine for point clouds (data points that represent geometric coordinates) to simplify the 3D modelling process. “Thanks to my experience as a student working on game development projects, I was used to developing algorithms and libraries that are fast, robust, and stable,” Xiao Qi says. “This turned out to be exactly the kind of work that dConstruct requires of their software engineers.”
Despite having strong computer science fundamentals, learning about the concepts that are a staple of robotics still took some getting used to. One such example is using simultaneous localization and mapping (or SLAM), which is a technological method used by engineers to build maps while determining where the robot or autonomous vehicle is within the map at the same time. This is done using advanced algorithms and input from sensory data. Now that he’s familiar with it, Xiao Qi has been tasked with researching different ways to make parts of their tech stack, including SLAM and robot autonomy algorithms, more robust.
According to the graduates, one of the things they enjoy most about their work at dConstruct is how well it aligns with their passion for computer graphics. “I enjoy working with native lower-level code and solving complex 3D problems, so the kind of solutions that dConstruct is working on in the field of robotics and graphics are a perfect fit,” Kah Wei says.
Brandon agrees, adding, “I feel lucky to have a job that is aligned with my goals.”
Projecting ahead, Brandon, Kah Wei, and Xiao Qi look forward to taking on more responsibilities and growing at dConstruct. They enjoy the challenges each project brings as they continue to learn new skills, dive into other disciplines, and ultimately become stronger graphics engineers.