IoT challenges for Smart Manufacturing
Download as PPTX, PDF•3 likes•4,204 views
1. The document discusses several challenges for connecting industrial IoT devices like a laser level transmitter to the cloud, including dealing with resource constraints, designing appropriate information models, standardization, security, and integrating autonomy. 2. It provides examples of industrial IoT applications and discusses using OPC UA as an information modeling and connectivity technology for the IIoT. 3. The presentation addresses how software architecture research can help with these challenges through reference architectures, patterns, modeling languages, performance modeling, and security methods.
![Overview
Internet-of-Things
April 30, 2018 Slide 3
Wikipedia Definition:
• “Network of physical devices, vehicles, home appliances and
other items embedded with electronics, software, sensors,
actuators, and connectivity which enables these objects to
connect and exchange data.”
Each thing uniquely identifiable and able to interoperate
within the existing Internet infrastructure
Direct integration of the physical world into computers
Improved efficiency, accuracy and reduced human intervention
IoT prerequisite for smart grids, virtual power plants,
smart homes, intelligent transportation and smart cities
Increased requirements towards privacy / security
Many application areas for Internet-connected devices!
[https://growthenabler.com/flipbook/pdf/IOT%20Report.pdf]](https://image.slidesharecdn.com/iotchallengesforsmartmanufacturing-180430182922/85/IoT-challenges-for-Smart-Manufacturing-3-320.jpg)
![Consumer Space
Internet-of-Things
April 30, 2018 Slide 8
Hardware:
• Texas Inst. DM3725CUS100 Digital Media Processor, 1Ghz
• Samsung K4X2G323PD-8GD8 256 MB LPDDR1 RAM
• SanDisk SDIN7DP2-4G 4 GB iNAND Ultra Flash Memory
Communication:
• Qualcomm Atheros QCA6234X-AM2D Wi-Fi and Bluetooth
Module
• Device Discovery
• REST APIs
• Smart Home Skill API
Powerful HW + REST-based communication
Amazon Echo Dot
[http://www.funbroad.tw/2017/06/alexa-skill-hello-world.html]](https://image.slidesharecdn.com/iotchallengesforsmartmanufacturing-180430182922/85/IoT-challenges-for-Smart-Manufacturing-8-320.jpg)
![OPC UA for Smart Manufacturing IIoT
April 30, 2018 Slide 10
[http://industrial.embedded-computing.com/articles/iic-connectivity-framework-defines-iiot-network-architecture-for-scalable-interoperability/]](https://image.slidesharecdn.com/iotchallengesforsmartmanufacturing-180430182922/85/IoT-challenges-for-Smart-Manufacturing-10-320.jpg)
IoT challenges for Smart Manufacturing
- 1. IOT-ASAP 2018 AT ICSA 2018, 2018-04-30 IoT Challenges for Smart Manufacturing Connecting a Laser Level Transmitter to the Cloud Heiko Koziolek, Senior Principal Scientist, ABB Corporate Research Germany
- 2. 1. Internet-of-Things Applications for Smart Manufacturing 2. IoT in Consumer Space vs. IoT in Industrial Space 3. 6 IoT Challenges for Smart Manufacturing 4. Software Architecture Research Challenges for IIoT Agenda
- 3. Overview Internet-of-Things April 30, 2018 Slide 3 Wikipedia Definition: • “Network of physical devices, vehicles, home appliances and other items embedded with electronics, software, sensors, actuators, and connectivity which enables these objects to connect and exchange data.” Each thing uniquely identifiable and able to interoperate within the existing Internet infrastructure Direct integration of the physical world into computers Improved efficiency, accuracy and reduced human intervention IoT prerequisite for smart grids, virtual power plants, smart homes, intelligent transportation and smart cities Increased requirements towards privacy / security Many application areas for Internet-connected devices! [https://growthenabler.com/flipbook/pdf/IOT%20Report.pdf]
- 4. Robotics Remote Service Industrial IoT Application Cases April 30, 2018 Slide 4
- 5. Electrical Vehicle Charging Industrial IoT Application Cases April 30, 2018 Slide 5
- 6. Container Ship Trim Optimization Industrial IoT Application Cases April 30, 2018 Slide 6
- 7. Condition Monitoring for Industrial Drives Industrial IoT Application Cases April 30, 2018 Slide 7
- 8. Consumer Space Internet-of-Things April 30, 2018 Slide 8 Hardware: • Texas Inst. DM3725CUS100 Digital Media Processor, 1Ghz • Samsung K4X2G323PD-8GD8 256 MB LPDDR1 RAM • SanDisk SDIN7DP2-4G 4 GB iNAND Ultra Flash Memory Communication: • Qualcomm Atheros QCA6234X-AM2D Wi-Fi and Bluetooth Module • Device Discovery • REST APIs • Smart Home Skill API Powerful HW + REST-based communication Amazon Echo Dot [http://www.funbroad.tw/2017/06/alexa-skill-hello-world.html]
- 9. Industrial Space Internet-of-Things April 30, 2018 Slide 9 Hardware: • ARM Cortex M4 Core: XMC4700-F144 Microcontroller based on ARM® Cortex®-M4 @ 144MHz • Flash: 2MB, RAM: 352KB Communication: • Today: 4-20mA analog HART communication • as IIoT Device: • Ethernet / IP • OPC UA Server for >200 device parameters • Connected to ABB Ability Edge Gateway & MS Azure Cloud Resource-constrained HW + OPC UA communication ABB Laser Level Transmitter LLT100
- 10. OPC UA for Smart Manufacturing IIoT April 30, 2018 Slide 10 [http://industrial.embedded-computing.com/articles/iic-connectivity-framework-defines-iiot-network-architecture-for-scalable-interoperability/]
- 11. OPC UA for Smart Manufacturing IIoT April 30, 2018 Slide 11 Laser Level Transmitter Communication: HTTP, UA Binary, AMQP, … Platforms: Windows, Linux, VxWorks, … Object-oriented Model: Static, dynamic, … Services: Connect, Subscribe, Alarms, History, … Features OPC UA recommended by Plattform Industrie 4.0, enables path to the IIoT/Cloud. Industrial Internet-of-Things Connectivity and Information Modeling
- 12. Challenge 1: Dealing with severe resource constraints Challenge 2: Designing appropriate information models Challenge 3: Implementing standards for interoperability Challenge 4: Bridging to cloud platforms Challenge 5: Addressing security requirements Challenge 6: Integrating autonomy, local analytics Challenge n: … How to connect a Laser-level Transmitter to the Cloud? IoT Challenges for Smart Manufacturing April 30, 2018 Slide 12 Let’s go through each challenge…
- 13. Memory Footprint for OPC UA Servers IoT Challenge 1: Resource Constraints April 30, 2018 Slide 13 Communication board on LLT has ARM-based Cortex M4 @ 144 Mhz, 2 MB Flash memory, 352 KB RAM The hardware to integrate is constrained by requirements for energy consumption due to explosion protection • e.g., power supply must be <570 mA (IEC 60079-27) • e.g., Raspberry Pi needs 2500 mA power supply ABB research project showed that the Unified Automation OPC UA High-Performance SDK can be squeezed into the target communication board Power consumption is currently at 641mW (for comparison Raspberry Pi 3: ~1200 mW), can be further optimized for 500mW target. IIoT connectivity must follow power/safety requirem. OPC UA on Laser Level Transmitter 0% 20% 40% 60% 80% 100% Flash: 845kB RAM: 35kB OPC UA Address Space OPC UA SDK embOS/IP embOS Device Firmware Others New software components for address space
- 14. Device properties, dynamic values, services IoT Challenge 2: Appropriate Information Models April 30, 2018 Slide 14 The OPC UA address space contains >200 variables (i.e., configuration parameters & sensor values) for the LLT, which is more than for a typical IoT consumer device Before IIoT, parameters encoded directly in proprietary firmware Migration path for LLT: 1. Define device parameters in Excel 2. Create XML NodeSet file 3. Transform to C-code using xml2c tool (UA address space) 4. Map information model to device firmware via API 5. Compile C-code into application This may be typical for many existing sensors and actuators. Providing appropriate content in OPC UA is non-trivial. Device Model on Laser Level Transmitter
- 15. Interoperability between devices of different vendors IoT Challenge 3: Standardization of Device Models April 30, 2018 Slide 15 OPC UA mainly provides a syntactic basis for interoperability Semantics of the device parameters are classically proprietary Devices from different vendors cannot be exchanged today Numerous standardizations ongoing: • NE131: ~30 standard parameters for sensors / actuators • OPC UA for Analyzer Devices (released) • OPC UA for Robotics (ongoing) • OPC UA for Industrial Drives (ongoing) • OPC UA for Process Automation Devices (planned) Metamodels for device models still emerging • IEC Common Data Dictionary being revised for functions Standardization could make IIoT devices exchangeable. NAMUR Core Parameters, OPC UA Companions
- 16. From devices to cloud analytics and apps IoT Challenge 4: Bridging to the Cloud April 30, 2018 Slide 16 Process control system with sensors and actuators provides a “trust boundary” and is normally shielded from Internet traffic IoT edge gateways collect data from one site and relay it to a cloud platform for analytics and access from mobile devices Edge gateway bridges between plant-internal traffic (e.g., OPC UA Binary) and plant-external traffic (e.g,. OPC UA via AMQP) Must be secured appropriately (auth., encryption, intrusion det.) Cloud providers supply edge gateway SDKs for vendors • Microsoft Azure IoT Suite • Amazon AWS IoT Core • Google Cloud IoT Core • … Security & Performance for IIoT gateways challenging. Edge Gateways
- 17. From isolated networks to denial-of-service attacks IoT Challenge 5: Security April 30, 2018 Slide 17 Laser level transmitter as IIoT devices provides much larger attack profile than for classical, proprietary devices LLT would be secured by OPC UA standard security mechanisms: user authentication, authorization, certificate exchange, encryption Further selected security aspects: • Risk Analysis, Network Segmentation, User Authentication, Secure Protocols, Secure Radio connections, Remote Maintenance, Intrusion Detection, Secure Product Life-Cycle, Component hardening, Cryptography, Isolation Techniques, … Standards: IEC 62443 series for “Industrial communication networks – Network and system security “ IIoT device & system security requires a comprehensive approach. Guidelines IIoT Security Machine Builder Guidelines for Industry 4.0 Security German Federal Office for Information Security: OPC UA Security Analysis
- 18. Making IIoT “really” smart IoT Challenge 6: AI/Autonomy April 30, 2018 Slide 18 Re-inforcement learning for Robots Intrusion/Anomaly detection for Process Control Systems Local Analytics / Production Optimization Skill Modeling for Discrete Manufacturing Automated Supply Chains Laser Level Transmitter: • Plug & Produce: could “learn/infer” appropriate configuration parameters for automated commissioning • cf. ICSA 2018 paper Limited “autonomous” devices today, but in research. Applying Artificial Intelligence & Machine Learning ABB YuMi Robot for Small Part Assembly A fully automated tendering process for personalized products sketched by Plattform Industrie 4.0
- 19. How to connect a Laser-level Transmitter to the Cloud? IoT Challenges for Smart Manufacturing April 30, 2018 Slide 19 1. Dealing with severe resource constraints 2. Designing appropriate information models 3. Implementing standards for interoperability 4. Bridging to cloud platforms 5. Addressing security requirements 6. Integrating autonomy, local analytics What can Software Architecture research contribute? Practitioners Challenges Reference Architectures Patterns & Decision Templates Modeling Languages & Transformations Performance Modeling & Benchmarking Security Methods Autonomy + Machine Learning SWA Research Topics
- 20. Patterns & Modeling Software Architecture & IoT Research Topics (1/2) April 30, 2018 Slide 20 Hammerstingl/Reinhard, ICIT 2015: Unified Plug&Produce Architecture Weyreich/Ebert, IEEE Software 2016: Reference architectures for the IoT Koziolek et. al ICSA 2018: Plug&Produce Reference Architecture with OPC UA Eloranta et al. Wiley Book 2014: Pattern language for distributed control systems Qanbari et al. IoTDI 2016: IoT Design Patterns Reinfurt/Leymann, EuroPLoP 2016: Internet of Things Patterns Thramboulidis et al. Computers in Industry 2016: UML4IoT Profile Eclipse Vorto: Device Description for IoT applications OPC UA Companion Specifications for Robots, Drives, Analyzer Devices Eclipse BaSyx Reference Architectures Patterns & Decisions Templates Modeling Lang. + Transformations
- 21. Quality of Service & Architecture Software Architecture & IoT Research Topics (2/2) April 30, 2018 Slide 21 deGooijer/Koziolek, ICPE 2012: Design Space Exploration for Robotics Backend Goldschmidt/Koziolek, IEEE Cloud 2014: Benchmarking Time-Series Databases Veichtlbauer et al., INDIN2017: OPC UA on resource constrained devices Dzung/Naedele, Proceedings IEEE, 2005: Security for Industrial Systems Wurm et al., ASP-DAC 2016: Security of Consumer and Industrial IoT Devices BSI 2017: OPC UA Security Analysis Pech, IEEE Software 2013, Software Agents in Industrial Automation Systems Vogel-Heuser 2014: Multi-agent based cyber-physical production system Musil et al. Springer Book 2017: Patterns for Self-Adaptation in CPS Eclipse 4DIAC Performance Security “Autonomy”
- 22. IoT Challenges for Smart Manufacturing Summary April 30, 2018 Slide 22 Many applications for the Industrial IoT: robotics service, EV charging, ships, … OPC UA as IIoT information modeling and connectivity technology SW Research Challenges: Patterns & QoS