Search and Rescue
Image: PixaBay 
Search and Rescue

Several companies around the world are building robots for Search and Rescue purposes. This is a great development, as several disasters have shown the need for such machines. However, those robots are still remotely managed by a human operator which limits their applicability severely. Disaster areas often have collapsed infrastructures and all kinds of problems that can interfere both with the possibility to come close enough for the operator and the required radio connection all together.

Another problem is that 'telepresence' which is the actual mode of operation, lacks the tactile feedback needed to do complex tasks. The DARPA Challenge has shown this in shocking detail. Some companies are working on better dexterity, but the operator being distant from the actual location will always be an impeding factor.


Full Autonomy is obviously a solution to those problems. But in this case, this autonomy need to be an integral part of the machine without any external need for information transfer. This is why we decided early in the ASTRID project that the system should run on very moderate hardware without any outside need for network connections or other forms of communication. It's obvious that such a machine, when data transmission is unavailable, couldn't carry a complete datacenter around.

Search and Rescue robotics for Disaster Relief where one of the primary targets of the ASTRID development. The ASTRID system is close to ready for implementation in such machines, and we are looking forward to work with the major players in this application area.

  External Articles
  • Rescue robots, the AI army that could save your life! 
  • One day, robots may work in zones too dangerous for humans 

© 2021-2022 MIND|CONSTRUCT & H.P. Willems  
Autonomous Reconnaissance
Autonomous Reconnaissance

Reconnaissance is critical in disaster stricken areas. If we can locate victims faster, their chances for survival rise.

Reconnaissance has the same problems with remote operators that Search and Rescue robots have. In addition, the limited availability of skilled operators limits the number of drones that can be up in the air after a calamity. Ultimately, we want to be able to launch whole swarms of intelligent reconnaissance drones that are capable of reacting to specific signs that might point to a victim in need of help. 


In this case, the fully independent mobility of the drone weighs in even harder as a prerequisite. The drones have to be light enough not to limit their action radius, and they should be able to survey as long as possible before needing any charging.

The ASTRID system can run on hardware as lightweight as a mobile smartphone. We configured our test setup very close to the processing and memory capacity of a smartphone, to be sure we can deliver on this requirement.

© 2021-2022 MIND|CONSTRUCT & H.P. Willems