Cognitive Computer Chess Comprehension or How to Play
- Slides: 16
Cognitive Computer Chess Comprehension or, How to Play Chess Against Something that’s Not Really There™
New Equipment Overview § SCORBOT-ER 9 § Industrial-Grade Robotic Arm § MVS-8000 § Frame Grabber and Embedded Vision System § Shiny new toys; what they do, and what we’ll do with them
SCORBOT-ER 9 § Robot Arm with 5 degrees of freedom § Can carry up to 2 kg (4. 4 lbs. ) § Position repeatability of 0. 09 mm (0. 004”) § Speed of movement defined in mm/sec or 1% of max. range
Control Software § Controller Internal Language, ACL (Advanced Control Language) § Interface to ACL controller provided via ATS interface software § SCORBASEpro — Windows-based software for control § Hand-held “teach pendant” allows direct manipulation of arm
Better Control Software § Robo. Cell — full § § compatability with SCORBASE programs Simulation software for arm and objects Direct interaction for position teaching 3 D graphical display for simulation Allows for offline manipulation and testing
Cognex MVS-8100 § PCI Frame Grabber & software visioning system § Captures 640 x 480 with external cameras § Included Cognex Vision Library(CVL) for capture/manipulation of images
Pat. Max § Most important piece of software included § Recognizes visual patters to identify objects § Can locate objects in a scene and give relative coordinates
Pat. Max Versatility § Fault tolerance in Pat. Max allows it amazing versatility in locating and recognizing objects § Despite surface irregularities and reflectivity, recognizes semiconductor wafer in varying conditions
Pat. Max Versatility, cont. § § § Identifies objects despite changes in scale Measures objects with accuracy within 0. 05% Successfully matches objects of random orientation
What to do with the system? § First, work on basic motions/pattern matching § Then, start putting things together to make a more complex system § Start with Chess!
Chess and the Robotic Arm § Goal — to be able to play chess against the arm § This will allow us to grasp the abilities of the system while leaning how to use it § Also, it’ll be really cool to be able to play chess against a robotic arm
Teaching Chess — Optics § Everything can be boiled down to pattern recognition § Recognition of each individual piece type/color § Recognition of board states — what changed?
Teaching Chess — Mechanical § Each time the arm has a move, it needs to accurately move a specific piece § Identify each square on the board by relative coordinate system § Need to measure distance to desired coordinates § Use well-known geometry to calculate how far to move in what direction § Capturing opponent’s pieces § Hit the clock (known location)
Teaching Chess — Fitness Algorithms § Many different kinds of fitness algorithms are around § Much discussion in Russell/Norvig § Marc and I consider the choice and construction of a fitness algorithm to be trivial § Left as an exercise to the audience members
Potential Pitfalls § Optics accurate enough? § Arm accurate enough? § Optical system — will it work? § Robot arm — will it be here?
Extra information § Robot Arm system description and specification at http: //www. intelitek. com/products/robotics/systems/scorb ot-er-9. html § Optical System specification at http: //www. cognex. com/marketing/products/prod_8000_ mvs 8100. asp § Ask Marc “EE” Marc
- Chess comprehension
- Biodiversity and classification
- Dumb kings play chess on fine grained sand
- Kings play chess on fat green stools
- Kings play chess on fat guys stomachs
- Elephants don t play chess summary
- Kings play chess on fine green silk
- Play chess with the boatman
- Cognitive and non cognitive religious language
- Swedish chess computer association
- Cbpt
- I've got a friend we like to play we play together
- Louise made the chocolate cake active or passive
- Play by play
- Hamlet kim
- Computer metaphor psychology
- Play go computer