A group of students from Purdue University has achieved a remarkable milestone by setting a new Guinness World Record for solving a Rubik’s Cube in just 0.103 seconds. This achievement eclipses the previous record time by approximately two-thirds. The innovative approach didn’t solely rely on speed; instead, an array of technologies including high-speed, low-resolution cameras, a specially designed cube for enhanced durability, and advanced solving techniques commonly utilized by human competitors played crucial roles.
The competition to build faster Rubik’s Cube-solving robots began back in 2014, marked by the success of Cubestormer 3, which utilized Lego Mindstorms components and a Samsung Galaxy S4 to solve the classic puzzle in just 3.253 seconds. The fastest time ever recorded for a human solver currently stands at 3.05 seconds, held by Xuanyi Geng. Over the past decade, engineers have successfully whittled down that record to mere fractions of a second.
Last May, Mitsubishi Electric engineers set a record with a robot that completed the puzzle in 0.305 seconds. However, the Purdue team—comprising Junpei Ota, Aden Hurd, Matthew Patrohay, and Alex Berta—broke this record by developing a robot they dubbed Purdubik’s Cube. Achieving a sub-half-second solution required a departure from Lego and the integration of optimized industrial components.
“Each previous record-holder has focused on one innovative aspect,” Patrohay stated in an interview with Technology News. In 2018, MIT students improved the technology by using superior industrial hardware, while Mitsubishi Electric upgraded to more suitable electric motors for this precision task.
One of the critical advancements from the Purdue team was enhancing the robot’s ability to rapidly visualize the scrambled cube. Unlike human speed cubers who observe the cube before timing begins, the robot’s clock starts as soon as it sees the cube. The students deployed two high-speed machine vision cameras from Flir, each capable of capturing three sides of the cube at once during brief exposures lasting as little as 10 microseconds.
Processing the data from camera sensors into a digital image takes time, which is why the Purdubik’s Cube employs a custom detection system that bypasses traditional image processing steps. This allows the system to focus on a small part of the sensor’s output—specifically, a cropped 128×124 pixel area—greatly enhancing data throughput.
Raw sensor data feeds directly into a rapid color detection mechanism, utilizing RGB measurements from minuscule sample areas on the cube’s squares, resulting in a swifter color identification process than methods incorporating AI.
“It may be slightly less reliable,” Patrohay explained, “but if it’s consistent at 90 percent, that’s sufficient for our needs given the emphasis on speed.”
While much of the robot’s hardware is custom-built, the team opted to use existing software for determining the most efficient solving method, leveraging Elias Frantar’s Rob-Twophase, which is a specialized algorithm designed for robots that can handle simultaneous spins of the cube’s sides.
The Robotic team also employed a technique known as corner cutting, which enables one side to start turning before another side has fully rotated. This overlapping movement reduces the time to complete the solution, providing a crucial advantage in their record pursuit.
While the corner-cutting technique offers speed benefits, it also presents risks; excessive force can lead to damaging the Rubik’s Cube during swift maneuvers. To navigate this challenge, the team had to finely tune both the timing of robot movements and the design of their customized cube.
Following precise guidelines set by the World Cube Association, the team modified their cube, ensuring it maintained the same functionality as a standard Rubik’s Cube while adhering to the requirements for color and texture. Enhancements included a stronger internal structure, crafted from custom 3D-printed SLS nylon.
The Purdue team’s design choice enclosed tight tensioning within the cube. “The cube we use for the record is tensioned incredibly tight,” Patrohay stated. “Turning it is extremely difficult and requires significant wrist action.” The tension, along with the use of lubricants, allows for better control at high speeds, enhancing the effectiveness of techniques like corner cutting.
The incorporation of faster servo motors contributes to improved solving times, albeit not without challenges. The Purdubik’s Cube utilizes six motors, each attached to metal shafts that connect to the center of each cube face. The team adopted a trapezoidal motion profile allowing the servos to accelerate rapidly—up to 12,000,000 degrees per second squared—while decelerating more gradually to maintain precise positioning as they halt.
Looking ahead, Patrohay sees potential for further enhancements to Purdubik’s Cube, suggesting that a cube constructed from a material stronger than plastic could improve performance and durability. “A specialized Rubik’s Cube crafted from carbon fiber composite could withstand higher speeds, leading to even quicker solving times,” he noted.
