Use case #1: Emergency/Disaster Recovery
When a natural or human disaster occurs, the first 72 hours are particularly critical to locate and rescue people. Nowadays, rescuers are assisted by different robots able to fly, climb or crawl, that are equipped with different sensors (e.g., temperature, gas, humidity, radiation, multi-spectral camera, heart rate, gyroscopes, compass, GPS, etc.) and wireless communication means. Victims are usually equipped with wearable sensors like the ones embedded in a cell phone (e.g., heart rate, camera, gyroscope) that could send useful complementary data.
NEPHELE will support the rescue team to:
- Locate and identify victims.
- Assess the victims’ injury.
- Assess the damages and comprehend remaining risks to prioritize the rescue operations.
Main technical challenges
Here this challenges rely on the heterogeneity of devices and time strong requirements. Data should be filtered and processed at different levels of the continuum to guarantee short delays while maintaining full knowledge of the situation. Devices are heterogeneous in terms of CPU, memory, sensors, and energy capacities. Some of the HW and SW components are very specific to the situation, while others are common to multiple scenarios. The network is dynamic because of links fluctuations, energy depletion of devices and device mobility. How to use VOs, where to deploy edge computing for what application in such a context is a tremendous challenge that NEPHELE can address. The orchestration of VOs and their performance are highly related to the hardware that manages them.
To mimic this use case, we will set a pilot using lab HW such as Turtlebots Niryo arms, Summit XLS with a UR5 arm with several depth cameras, lidars and sensors.
Use case #2: AI-assisted Logistics Operations in the Port of Koper
Port of Koper is a multi-purpose deep-sea freight port and the most important container port in the Northern Adriatic Sea, with a total maritime throughput over 19,5 million tons in 2020. Port terminals are equipped with state-of-the-art transshipment and warehousing equipment, such as ship-to-shore cranes, reach-stackers, forklifts, utility tractor rigs, etc. Port of Koper has excellent connections to road and railway network. Continuous monitoring and optimization of the traffic within the port poses a daily challenge (traffic congestions, unplanned road closures, etc.).
NEPHELE will optimize one of the most important operations in the port: the routing of containers from the container terminal yard or Depo area to different Container Freight Stations (CFS), where the cargo is stuffed/stripped, and viceversa. This will:
- Help to reduce routing times.
- Get lower CO2 emissions.
- Get higher truck/forklift utilization and service level agreements (e.g., times of delivery, compliance with goods sensitivity, etc.).
- Allow to exchange and aggregate data among the physical components involved in the use case (e.g., forklifts, trucks) through VOStack layers.
- Comply with security and low latency requirements thanks to the application of decentralized machine learning techniques at a VO level.
- Allow the orchestration of the deployed microservices between the cloud and edge computing orchestration platforms - ensure the self-healing, portability and elasticity of the whole application graph.
Main technical challenges
- Container localization.
- Truck/forklift availability and location information.
- Route status aspects.
- Information on policies and KPIs that need to be achieved.
- Algorithms for the real-time allocation of resources and for the routing of containers from the container terminal to CFS, where the cargo is stuffed/stripped, and viceversa.
- AI-assisted Logistics Operations with continuous monitoring of the status of devices involved in the port logistics operations and failure alerts or any obstacle occurring during the process (e.g., traffic congestions, road closures, railway crossings, etc.).
The IoT devices in this pilot will include industry-grade 5G IoT gateway with additional computing capabilities, industry-grade UHD cameras, 5G body-worn cameras and various 5G mobile devices/sensors for business process support.
Use case #3: Energy management in smart buildings/cities
Smart applications and services leveraging on the VOStack can help to manage control actions of building equipment, providing user with customized services for energy-efficient, well-being and comfort. This is implemented by deploying an automation scheme that gathers real-time information from a variety of IoT devices –appliances, sensors, HVAC– together with Edge nodes that will instantiate VOs. This approach avoids bottlenecks caused by placing all the intelligence in a centralized Smart Building/City monitoring and control system.
NEPHELE can help to reach a high-performance level of benefits in terms of:
- Latency.
- Energy consumption.
- Reliability of the offered services.
Main technical challenges
Challenges for the distribution of smart energy management applications into different and interoperable components in the cloud-edge continuum are related with heterogeneous performance requirements:
- Energy Efficiency Management security, which is of paramount importance for smart building applications, especially when end users’ data form part of the decisions processes.
- Real-time or almost real-time execution of analysis pipelines, since it directly affects the impact of the complex event decisions that are applied constantly at a microservice level.
Special equipment: the existing IoT and Edge/Cloud computing infrastructure is mainly composed by wireless microcontroller-based system-on-chip devices communicating with edge nodes. The IoT constrained devices include electromechanical sensors and actuators, which monitor and control different environment parameters. In turn, these devices communicate with edge devices through IoT communication protocols such as CoAP or MQTT, over low-power radio technologies such as LoRaWAN, Zwave, 6LoWPAN, or NB-IoT.
Use case #4: Remote healthcare services
The current ultrasound medical imaging processes are constrained by both the technical features of the local device and the knowledge of the local healthcare operator performing the examination. In fact, Electronic Health Record (EHR) processes are currently bound to on-premises dedicated hardware/firmware components to fulfil the need of a real-time or an almost real-time execution of the process. As a result, acquisition expenses are very high and limit the degrees of flexibility in upgrading the hardware and, consequently, the types and number of functions that can be locally provided.
The goal of this use case is to connect, and somehow to decompose and virtualize ultrasound medical imaging systems into the cloud-edge continuum to lose any barriers due to the hardware capabilities and localization of current physical systems. Functions refer to those EHR processes that elaborate ultrasound data to provide the operator with additional qualitative information (often visualized over colored overlay images over the black and white video) or quantitative data (spatial measures, pattern identifications, etc.).
NEPHELE will:
- Allow to exchange data and resources among the physical components involved in the use case (e.g., acquisition hardware, monitors, potentially other interactive HW input devices such as keyboards).
- Provide additional capabilities such as distributed data management and analysis based on machine learning (ML) techniques, authorization, security and trust based on security protocols and blockchain mechanisms, etc.
- Allow the orchestration of data and resources between the cloud and edge computing orchestration platforms required for the proper and dynamic interplay of the functions in the application graph thanks to the integrated meta-orchestration framework.
Main technical challenges
- Heterogeneous performance levels required by the different functions (falling from “Tactile Internet” requirements to ones generally provided by current cloud systems), and to the way data is treated.
- Data security, which is of paramount importance for medical processes and so is the need of a real –or almost real– time execution of the process.
Ultrasound acquisition hardware and the medical imaging viewers will be required for this use case.