Drone operations are characterized by low cost, speed, safe access to dangerous areas and flexible and scalable deployments.
When a catastrophe occurs, autonomous drones can be used to collect data in place efficiently and transmit it through different communication networks.
In the 5G-DIVE project we intend to improve the current navigation systems to allow the local treatment of this information and the control of the drone’s trajectory, so that it can be modified dynamically.
5G-DIVE is a research project belonging to the second phase of the collaboration framework between the European Commission and Taiwan for the development of 5G technologies.
This initiative, which we coordinate from the Carlos III University of Madrid , aims to create a platform for the provision of services with low latency and optimized thanks to artificial intelligence.
In the case of drones, the aircraft will be integrated into the 5G-DIVE platform, where they will process the collected data locally. This allows information to be managed autonomously.
5G connectivity will allow the execution of artificial intelligence algorithms on the network, too expensive to run on drones. In this way, we endow them with superior intelligence. An intelligence that is distributed and separated from drones, which are limited to executing orders commanded by the network.
This intelligence also enables them to process images. A very useful possibility in the event of a catastrophe and that allows:
- Integrate several 2D images to create a continuous map of the affected area.
- Recognize patterns of certain events, such as fire in buildings or risks to human lives.
The 5G-DIVE project
The objective of the project is to build an integrated computer network using computers and other devices close to users (what is known as fog computing or fog computing ), small data centers deployed by the operator and public clouds .
The benefits of this design include:
Low latency, due to the fact that the computing units are close to the user (unlike remote servers).
- The characteristic scalability of highly distributed systems.
- Opportunity to create new business models thanks to the use of computer and network resources distributed in the same place where the service is provided.
- A shopping center could deploy computer nodes (computers, mobiles, etc.) and rent them for the development of third-party applications. For example, a service for reloading vending machines through a robotic platform or the deployment of services to manage sales and banking in the mall without the need for external infrastructure.
But the project is not limited to the creation of this infrastructure. It extends it with artificial intelligence and automation capabilities so that infrastructure management is as agile as possible.
Applications in the industry
One of the objectives of the 5G-DIVE platform is to support industries in daily operations, management and automation of business processes.
A vertical industry, such as the use of drones for logistics, allows automating network management. You can ask the network for a series of intelligent services based on data collected in situ and inference or learning algorithms that the same network will train and provide to the vertical industry.
One of the pillars of our project is the validation of 5G technology both for aircraft handling and for the control of robots used in industrial facilities and factories. For this, a series of pilots will be carried out in the offices of two vertical industries in Taiwan.
Digital twin applications
A digital twin is a software copy of an industrial robot. Share sensor and control data, so that any robot behavior can be replicated in real time in the digital model and vice versa.
The project aims to demonstrate 5G performance in the interconnection of a real robot with its digital twin, sharing computing resources and software with the robot.
The platform will provide all the necessary functions to control the robot in the virtual infrastructure and to predict the movements of the robot through artificial intelligence.
The 5G network coverage will allow real-time control and a remote view of the robot’s status and performance without having to operate directly on the physical machine.
Error detection in an assembly line
Assembly lines in the future will be operated entirely by robots, which will be controlled by a central cloud or, in this case, by a computing platform.
The robots will not execute complex computer routines to manage their activities, but all control algorithms will be executed on a centralized platform.
This allows a process of optimization and updating of robots much faster, as well as possibilities of coordination between key robots for the future development of the industry.
An interesting case within this vision is the development of assembly lines that are capable of monitoring products and identifying those that contain a defect. One of the 5G-DIVE pilots is based on the detection in the network, using artificial intelligence, of products that do not meet a quality standard.
It is important to emphasize that algorithms run on the network. The assembly line will send videos of the elements that pass through it and the network will execute detection algorithms that cannot be executed on the assembly line itself.
A key role in this pilot is 5G technology. Without low latency in the network, it is not possible to perform a time detection that allows acting on the defective product.
Press Beth Daley Editor and General Manager February 5, 2020 6.13pm EST