IMOCO4.E developments are grouped in ten building blocks (BB) which have been integrated into an edge-to-cloud open platform for intelligent control and optimized lifecycle of industrial mechatronic systems. The platform has been deployed on commercial HW developed by IMOCO4.E partners and interoperability with commercial industrial robots (fixed, modular) and legacy systems has been proven. This is demonstrated and validated in five pilots, four demonstrators and four use cases:
BB1: SoC/FPGA platforms for smart control and signal processing. Workflow tools for FPGA configuration and validation, compatible with multiple vendors. Network configurations for up to 1000 Mb/s data rates using wired or wireless networks. Tools for network configuration management and latency optimization and reliable TSN.
BB2: High speed Vision in the Loop. High-speed image capture, achieving real-time image processing and high-performance computing within a single embedded system.
BB3: Novel sensors. Embedded algorithms for Mbps communication over motion drive cables with novel EMI suppression techniques and degradation detection.
BB4: Real-Time smart control platform. Embedded platform with reusable HW/FW/SW IP, and advanced multi-source parallel data processing and communication.
BB5: Smart control algorithms library. Advanced (de)centralized control strategies for multivariable systems, leveraging high-fidelity models and AI-based tuning extensions to adapt to dynamic disturbances and optimize control systems.
BB6: Algorithms for condition monitoring, predictive maintenance and self-commissioning of industrial motion control systems. Predictive maintenance functionalities through AI techniques and digital twins, enabling virtual commissioning.
BB7: High performance servo-drives. Scalable servo drive platform for low-power DC drives with FW supporting configuration and control of rich IO connectivity.
BB8: AI-based components. AI-accelerator SW/HW implementation in FPGA prototypes and ASIC, for parallel and determined execution. Demonstrated proof-of-concept for pick-and-place systems capable of transferring learned operations from virtual to physical environments.
BB9: Cyber-security tools and trustworthy data management. AI-based anomaly detection and real-time event management systems for IIoT environments, conducting thorough vulnerability assessments to ensure system security.
BB10: Motion / path planning, collision avoidance and navigation algorithms. Multi-sensor navigation algorithms enabling robust robot navigation in indoor environments. Enhanced real-time motion planning algorithms and reinforcement learning approaches under laboratory conditions for autonomous robotic manipulators.
Pilot 1: Tissector. An automatic medical device in the field of digital pathology with integrated automatic calibration, verification, and partial self-commissioning. Extended remote monitoring, AI-based predictive maintenance, and failure mode recognition is enabled.
Pilot 2: Semiconductor production. A high-speed and high accuracy die bonding platform and semiconductor production line with 10% process improvement is demonstrated. Machine-to-machine performance calibration is eliminated via in-line feedforward tuning.
Pilot 3: High speed packaging. In-line filling & stoppering machine, Tea bag machine for cotton thread knot technology, achieving improvements in +5% system efficiency, 80% data utilization, and 20% reduction in resource consumption. TSN for interoperability and a fully distributed microservice architecture for edge/cloud intelligence sharing was implemented.
Pilot 4: Healthcare robotics. X-ray intervention medical manipulator with digital twin models for monitoring, control, and condition assessment, identifying mechanical wear and electrical faults.
Pilot 5: Mining / tunnelling robotic boom manipulator with collision avoidance and advanced path planning with high success in movement requests.
Demonstrator 1: Consumer goods. High precision cold forming of complex 3D metal parts for shaving devices (multi-stage line) with reduced downtime or faulty products, reduction in ramp-up time, and decreased scrap rates. A lightweight digital twin technology for model-based control and predictive maintenance was implemented.
Demonstrator 2: Injection molding. Embedded RFID in injected plastic parts and low-powered wireless technology for part traceability and real-time parameter monitoring.
Demonstrator 3: Warehouse logistics. Autonomous intra-logistics transportation demonstrated with self-organizing, autonomous transport capabilities, improved fast obstacle avoidance and navigation.
Demonstrator 4: Cosmetic production line. Vision-based (AI) improved pick & place robotics for randomly arranged and differently shaped bottles with high reliability.
Use case 1: Industrial drive for smart mechatronics applications. Demonstrated a system for identifying Remaining Useful Life (RUL) and anomalies, improving maintenance planning and operational reliability.
Use case 2: CNC for integrated machine tool and robot control. Enabled simultaneous control of machine tool and robot, self-commissioning of control loops, and real-time adaptation with AI techniques.
Use case 3: Tactile Robot Teleoperation. Remote teleoperation and virtual reality modelling digital twin for haptic teleoperation (hand gesture recognition). Implementation of behavioural prediction solutions on SoC+FPGA platform, accounting for latencies and deficiencies.
Use case 4: Advanced and Intuitive robot control and programming. Methodology and algorithms for intuitive robot Human-Machine Interface (HMI) with advanced trajectory planning for operation in complex environments. Integration of the robot digital twin to robot programming process.
