How Jigsaw Works: The Puzzle-Solving Robot

Introduction

• Jigsaw: A specialized robot designed to solve jigsaw puzzles extremely quickly.
• Development timeline: 3 years of development.
• Performance: Initial tests indicate it is 200 times faster than the fastest human competitive jigsaw puzzler.

Human Puzzle-Solving Abilities

Key Steps in Puzzle Solving

1. Picking up a piece:
   • Involves hands with 27 bones, 34 muscles, and a high concentration of nerves.
   • Allows for precise and dexterous movements.
2. Rotating the piece:
   • Requires fine motor skills to orient the piece correctly.
3. Moving the piece into position:
   • Utilizes the entire arm, showcasing complex bone and muscle coordination.
4. Deciding where the piece should go:
   • Mixes visual perception, pattern recognition, spatial reasoning, and executive function.
   • Human brains process these elements quickly and subconsciously.

Unique Human Attributes

• Despite limited physical prowess in other areas, humans excel due to their complex brains.
• Our brains consume 20% of our daily energy, which supports advanced functions like tool use, problem-solving, and language.

Developing Jigsaw

Challenges

• Emulating the four advanced steps humans naturally perform took extensive research and trials.

Solutions for Each Human Ability

1. Picking Up Pieces:
   • Utilized a specialized suction cup commonly used on assembly lines.
   • Equipped with a solenoid and vacuum pump for precise control.
2. Rotating Pieces:
   • Suction cup grabber attached to a finely tuned donut motor.
   • Offers precision down to 0.005 degrees.
3. Moving Pieces:
   • Modified an avid CNC router with ClearPath industrial servo motors.
   • Provides accuracy to 0.0005 inches.
4. Deciding Piece Placement:
   • Overcame the hardest problem by initially struggling with pattern recognition and spatial reasoning.
   • Shifted approach to edge analysis rather than visual pattern matching.

Edge Analysis Method

1. Image Collection:
   • Utilize a cell phone camera to take pictures of each piece.
2. Spline Matching:
   • Convert each piece's edges into four splines.
   • Match spline lengths to reduce solution space.
   • Calculate overlapping area to rank potential matches.
3. Solution Space Mapping:
   • Jigsaw starts with corner pieces and maps possible solutions, adjusting for mismatches.

Scaling to a 1000-piece Puzzle

• Testing: Successfully solved smaller puzzles; aimed for a larger challenge.
• Challenges: Issues like accumulated error from piece misalignment.

Final Adjustments

• Spring-loaded linear slider for feedback: Simulates human tactile feedback.
• Software enhancements: Implement routines for precise placement and correction.

Robot vs. Human Face-Off

Preliminary Contests

• 30-piece puzzle: Tammy McLeod, world record holder, demonstrated rapid solving.
• 500-piece puzzle: Faced off against actress Kristen Bell with Tammy's guidance.

Ultimate Showdown

• 1000-piece puzzle: Jigsaw vs. Tammy McLeod.
• Performance: Jigsaw evaluated and placed pieces efficiently, using new upgrades to correct placements.
• Outcome: Jigsaw successfully completed the puzzle, showcasing robot advantage in precision and endurance.

Additional Notes

• HackPack Introduction: For teenagers and adults, offering programmable robots and engineering skills training.
• Human Puzzling Tips: Tammy McLeod shared four expert tips to improve speed and efficiency in solving puzzles.

Conclusion

• Jigsaw successfully surpassed human capabilities in solving complex puzzles.
• Demonstrated the effectiveness of specialized robotic systems over human abilities in specific tasks.
• Potential for wider applications and enhanced tinkering through educational packages like HackPack.