What could autonomous ground-based systems help us solve?

Many inspection, monitoring and security tasks need to be carried out in environments where drones are restricted, difficult to use or simply not the right tool. In other situations, sending people may be repetitive, inefficient or unsafe. 

A ground-based autonomous system could offer another way to operate in these environments, moving close to equipment and infrastructure while carrying cameras and sensors and using AI to interpret what it encounters. 

We talked to researcher Richa Upadhyay to learn more about the technology behind this rover demo, what it is designed to explore, and where it could make a difference in practice. 

Developed within Digital Testbed Sweden (DTSE), the rover combines autonomous mobility, AI-based visual understanding and edge computing over a standalone 5G network. 

Watch the demo video on LinkedIn

In simple terms, what are we seeing in this demo? 

The rover is being developed as a platform for demonstrating autonomous mobile systems. It is intended to use computer vision and vision-language models to understand its surroundings, identify objects or situations, and make decisions with limited or no human control. 

Some AI processing can run directly on the rover. For more demanding tasks, it can connect through a standalone 5G network to nearby network edge infrastructure, where heavier computations can be handled with lower delay than if everything were sent to a distant cloud. 

What new capability does this unlock? 

The key capability being explored is the combination of autonomous mobility, AI-based visual understanding and flexible computing. 

Simple or time-critical tasks can be processed directly on the rover, while more complex AI tasks can be offloaded to the network edge through the standalone 5G network. 

Network edge computing can offer several advantages: 

• Lower latency, compared with sending data to a distant cloud. 
• More computing power than the rover may have available onboard. 
• Better energy efficiency, since the rover does not need to carry powerful computing hardware for every task. 
• More flexible AI deployment, allowing advanced models such as vision-language models to run and be updated at the edge. 
• Improved scalability, since several rovers or connected devices could share edge resources rather than each requiring high-end onboard hardware. 

Where could this have the biggest impact? 

This type of rover could be useful wherever automated ground-based monitoring or inspection is valuable, particularly where drones are restricted or where human access is difficult, repetitive or unsafe. 

Possible applications include: 

1. Security patrols at industrial sites, campuses, warehouses or restricted areas. 
2. Infrastructure and asset inspection, such as checking equipment, cables, pipelines, telecom sites or utility areas. 
3. Field monitoring at construction sites, energy facilities, agricultural environments or outdoor test sites. 
4. Industrial automation, supporting routine checks, inventory observation or anomaly detection. 
5. Emergency and safety support, for example entering an area that may be unsafe for people before personnel are sent in. 

From research to something you can see and test 

At RISE, we continuously prototype, test and combine technologies to explore how they could work in practice. 

Prototypes like this rover allow researchers to investigate new possibilities, uncover limitations and understand what is needed to move emerging technologies closer to real-world use. 

They also give us a way to make our research more tangible and accessible, by showing how emerging technologies can be explored and tested in practice. 

Intrigued? There’s more to explore. 

Explore the connected projects and research behind the demo, and make sure to subscribe to the AI at RISE newsletter for more examples of AI research in action. 

Want to get into the technical details? Contact researcher Richa Upadhyay at richa.upadhyay@ri.se. 

Connected projects: 

1. Reality-based test environments drive digital innovation in collaboration 
2. ICE Datacentre - Edge and 5G  
3. NEXSoS paves the way for more efficient data sharing through new standards