In the fall of 2021 RAINBOW is finalizing the early release of the three project demonstrators. These are virtual demonstrators for the Human Robot Collaboration, Digital Transformation of Urban Mobility and Power Line Surveillance use cases, to evaluate the first version of the RAINBOW platform.
RAINBOW shall be deployed at the premises of each of the three project demonstrators with RAINBOW components installed on use case specific fog devices (e.g. Raspberry Pi’s, Jetson Xavier, etc.). Use case specific applications will be integrated with those components and the capabilities offered to the demonstrators shall be enriched:
∙ Access to the Service Graph Editor, a UI used to create an application graph, which represents the interfaces between the application components and their dependencies.
∙ Use of RAINBOW’s Monitoring Agent, which collects data at application level and at resource utilization level from the deployed containers and the fog devices, with the help of Monitoring Probes.
∙ Definition of the Service Level Objectives, which ensure that the system remains in optimal (or near-optimal) condition for performing its duties, while its topology and number of nodes change.
The demonstrators will put their use cases to the test. They will analyse the platform’s initial capabilities and start collecting measurements with RAINBOW, based on already defined metrics and KPIs. Those measurements shall be then compared to the baseline values, in other words the values of the previous set-up (before RAINBOW); they will be also available to be compared with the values of the more mature 2nd version of RAINBOW in Q3 2022.
A few words about the RAINBOW tests on each demonstrator:
A Human Robot Collaboration system is a collision prediction and avoidance system aimed at reducing risk of accidents involving personnel and robots in an indoor work environment. By collecting personnel and robot 3D coordinates and motion dynamics at periodic intervals, predictions on collisions are made. With RAINBOW, it will demonstrate: (i) a decrease in system latency for collision prediction and avoidance, (ii) reliable hand-off of 3D coordinates and motion dynamics data between monitoring nodes and (iii) scaling-up and scaling-down of application services based on the number of robots and personal in the same work environment (which can dynamically change). RAINBOW shall also offer support services like data sharing with registered users for monitoring and analytics and attestation of new devices which need to be onboarded in the work environment.
In the Digital Transformation of Urban Mobility use case, a real-time geo-referenced notification system is created to dispatch alerts to vehicles moving in an area where a hazardous event is detected. Those alerts can be sent directly to a vehicle from a road side unit, if the vehicle is close to the hazardous event, or disseminated via a cloud node, with low priority, to other vehicles which are farther away from the hazardous event. With RAINBOW, it will demonstrate: (i) the orchestration of the Automatic Hazardous Events Detection service migration between the cloud node and the road side unit in order to keep bandwidth occupancy and energy consumption within a predefined value range, (ii) a decrease in system latency and an increase in message throughput (i.e. number of C-V2X events managed), and (iii) how to demodulate those messages and draw the hazardous area on an online map, then offered to everyone via the use of a mobile app.
In the Power Line Surveillance use case, a swarm of drones is used to inspect critical infrastructure in areas with difficult or no accessibility. Those drones operate in a constantly changing environment, e.g. changing weather conditions and especially wind power and direction, the topography of the terrain, interference from local electromagnetic fields of power lines or other sources of electromagnetic waves, the nature of the inspected object and even the accessibility of the area to deploy system nodes. This demonstrator will implement a distributed Ground Control System that will govern a swarm of drones to optimize their operations and increase the swarm’s range and autonomy. With RAINBOW, it will demonstrate (i) a decrease in time to pass the control over the drone from one Ground Control System to another, (ii) a decrease in data acquisition time per Kilometer of power line, (iii) a reduction of overlaps between individual flight routes, and (iv) better efficiency of battery usage for a productive phase of the drone flight.
