Gregorio Orlando

Abstract— Desk-based learning and creative activities benefit from handwritten engagement. However, current generative AI tools deliver guidance through a separate screen, creating a gap between where users think and where assistance appears. To address this, in this work we design AIfred, a desk-based robotic arm with a projector mounted at the end-effector that places AI-generated guidance alongside handwritten work. AIfred combines workspace perception, context-aware content generation, and robot-mediated projection to support math- assignments, image-generation, and drawing tasks. In a user study (n = 36), we compared AIfred against ChatGPT (GPT-5.6 Luna) running on a laptop. Both tools performed comparably while assistance was available during math assignment (6.7 vs. 7.3/10, p = .41), but AIfred improved short-term learning transfer by 60% once assistance was withdrawn (7.0 vs. 4.4/10, p = .003). In addition, independent art and design professors ranked drawings produced with AIfred better in 33 of 36 cases. Our findings indicate that spatially co-located AI assistance benefits tasks whose guidance shares a spatial frame with the work.

Submitted to ICRA 2027, under review.

Quadrupeds are easy with harmonic drives and torque sensing, and hard with hobby servos. Botzo is the hard version, on purpose: printed parts, 25 kg servos and a Raspberry Pi, documented so other people can rebuild it. Our community is already doing so, and the Botzo repo has been forked 4 times, and starred 13 times.

Low-inertia leg geometry, closed-form inverse kinematics, gait planning, and ROS 2 Jazzy orchestration. The current work replaces that hand-tuned gait with a policy trained in NVIDIA Isaac Lab.

A LiDAR SLAM map is millions of triangles: ideal for a robot, hostile to a HoloLens that must also render an application at frame rate. Most of that detail is irrelevant, if the XR scene only needs to know where the walls are, surface roughness is pure cost.

Six months at the SnT Automation & Robotics Group, on the conversion step: extract the planar wall structure, discard the rest, hand Unity something it can draw. 4.27 M triangles to 18,094.

Presented at the VR Summit 2026 in Germany, and used in Human Interaction for Collaborative Semantic SLAM using Extended Reality.

No weapon, by choice: a low polycarbonate wedge on four-wheel drive, symmetric about its horizontal plane so being flipped changes nothing. An ESP32 takes a PS4 controller over Bluetooth and drives four motors through an L293N.

Fourth of eleven on the team's first attempt, out in the semi-final when the chassis cracked. Plastic wheels shattered, the motors could not out-push anyone.