Successfully reported this slideshow.
Design and Implementation ofArtificial Neural Networks in Semi-autonomous Robotic SystemJohn Lau Kah Soon (7D3A1916 / 1275...
Summary• Problems Identification• Methodology• Solutions & Achievements• Solutions & Achievements• Conclusions & Future Di...
Problems IdentificationProblemsHow to build a semi-autonomous robotic system ?Most ANN implementation SoftwareCan ANN be i...
MethodologyPseudocodeMATLABHardware Setup[1] Boe-bot (BS2, BoE)[2] eb500Java (SE 5.0)[1] Communication APIVHDLStartMicroco...
Heimdrix Type 0 SystemBasic Stamp 2(Heimdrix.bs2)MobilitySubsystemIR ObjectDetectionSubsystem[2], [3]WirelessCommunication...
Definition BACKEND
RICMM BACK0.0081 0.9575 0 1 0000 1111ˆ0.9339 0.0198 1 0 1111 0000− −     = ≈ =     − − −     MRICMM CMM ...
ANN Controller Chip BACKEND
Conclusions & Future DirectionConclusionsANN + RoboticsConcurrencyRICMM 6 gatesXOR, RandomSoftware ANNHardware ANNEfficien...
The EndQ & A Session123456789Cover PageSummaryMethodologyHeimdrix Type 0 SystemProblem IdentificationDefinitionRICMMANN Co...
Upcoming SlideShare
Loading in …5
×

Design and Implementation of Artificial Neural Networks in Semi-Autonomous Robotic System

705 views

Published on

Published in: Technology
  • Be the first to comment

  • Be the first to like this

Design and Implementation of Artificial Neural Networks in Semi-Autonomous Robotic System

  1. 1. Design and Implementation ofArtificial Neural Networks in Semi-autonomous Robotic SystemJohn Lau Kah Soon (7D3A1916 / 12753189)Supervisor: Terence Tan END
  2. 2. Summary• Problems Identification• Methodology• Solutions & Achievements• Solutions & Achievements• Conclusions & Future DirectionEND
  3. 3. Problems IdentificationProblemsHow to build a semi-autonomous robotic system ?Most ANN implementation SoftwareCan ANN be implemented in Hardware ? How ?ProblemsHow to design ANN for robotics application ?How to embed ANN into semi-autonomous robotic system ?Embedded system  General-purpose computer ?END
  4. 4. MethodologyPseudocodeMATLABHardware Setup[1] Boe-bot (BS2, BoE)[2] eb500Java (SE 5.0)[1] Communication APIVHDLStartMicrocontrollerTechnology[1] Communication API[2] FreeTTS API[3] StandardSerialPort.java[4] Utils.javaPBASICANN Controller Chip onXilinx Spartan-3E boardVHDLSchematicsHeimdrix.bs2, Heimdrix_Type_0.jarHeimdrix Type 0 SystemFuture R&DPLD (FPGA)TechnologyUDMIncrementalProcessModelEND
  5. 5. Heimdrix Type 0 SystemBasic Stamp 2(Heimdrix.bs2)MobilitySubsystemIR ObjectDetectionSubsystem[2], [3]WirelessCommunicationSubsystem [1]BluetoothUSBAlarmSubsystemEmbeddedSystemCMMPerceptronRICMMRICMMIR ObjectDetectionSubsystemMobilitySubsystemNeural NetworksSubsystem[a], [c], [d][a], [d]SOFTWAREANNHARDWAREANNConclusionKeyboardUSBDongleGeneral-purposeComputer(Heimdrix_Type_0.jar)Heimdrix.javaNeuralNetworks.javaStandardSerialPort.javaUtils.java Keyboard.javaCMM.vhdCustomCustom--mademadeANNANNControllerControllerChipChip‘3’ ‘7’[1] [2] [3]Telemetry[a] Wireless Comm. Subsystem (Software)[a] Wireless Comm. Subsystem (Software)[b] Concurrent GUI[b] Concurrent GUI--based Command Subsystembased Command Subsystem[c] Error Detection Subsystem[c] Error Detection Subsystem[d] Self[d] Self--repair Subsystemrepair Subsystem[e] Speech Subsystem[e] Speech SubsystemAssociation withSemi-autonomousRobotic System[b][a], [b], [c], [e]Logging &ConversionpurposesEND
  6. 6. Definition BACKEND
  7. 7. RICMM BACK0.0081 0.9575 0 1 0000 1111ˆ0.9339 0.0198 1 0 1111 0000− −     = ≈ =     − − −     MRICMM CMM Hardware representationEND
  8. 8. ANN Controller Chip BACKEND
  9. 9. Conclusions & Future DirectionConclusionsANN + RoboticsConcurrencyRICMM 6 gatesXOR, RandomSoftware ANNHardware ANNEfficientMethodologyRigorousENDFutureDirectionANN ScalabilityRigorousMathematicalAnalysisMore SensorsAnd ServosUpgradeFuzzy LogicGenetic AlgorithmSpace ExplorationSpace ExplorationRescue OperationsRescue OperationsOil & Gas IndustryOil & Gas IndustryMining IndustryMining IndustryServicesServicesSecuritySecurity
  10. 10. The EndQ & A Session123456789Cover PageSummaryMethodologyHeimdrix Type 0 SystemProblem IdentificationDefinitionRICMMANN Controller ChipConclusions & Future Direction

×