Mobile Sensing for Safety

The research project 'Mobile Sensing for Safety' (MoSeS) is aimed at designing a system that supports firebrigades in performing their task to control fires in complex buildings. Availability of adequate information for firefighters and commander is crucial for the safety of the firefighters and the effectiveness of their operation. Nowadays very little information about the firefighters in a building is available to the commander, who generally remains outside the building. For example, he does not know where exactly his men are in the building. Nor does he know how well they are. Furthermore, he does not know the compressed air level of the firefighters. So he has no clue how long they can stay in the building and whether they are in danger.

The purpose of the MoSeS system is to provide real-time situational awareness to the commander. Using the MoSeS system, the commander can see the exact location of his firefighters on the floorplan of the building, and he can see a prediction of the health risks (such as heat stroke, exhaustion, etc) of each firefighter. Also he can see on the floorplan where the location of the fire, potential casualties and dangerous substances are. Having this real-time situational awareness, he is able to make better informed decisions.

The MoSeS system is built around modern mobile technology, which has been chosen for its fantastic sensing, communication, processing and visualization capabilities, yet at acceptable cost. Firefighters are equipped with a commercial-off-the-shelf smartphone, which collects information from sensors that is carried in the equipment via Bluetooth. Typical sensors are inertial sensors (acceleration, compass, etc) for positioning purposes and physiological sensors (heartrate, respiration, temperature) for health monitoring. In the smartphones this bare sensor data is processed to become relevant information. Smartphones mutually set up a mobile ad-hoc network (MANET), thereby being independent on existing wireless network infrastructure like 3/4G and WiFi access points. Notice that in complex buildings mobile network reception is not always good, and that firebrigades turn off power, and thus access points, when they enter a building. The sensor data is sent hop-by-hop over the MANET to the tablet of the commander, which is also part of the MANET. At this tablet, the data is further filtered and processed and then visualized at the floorplan of the building.

The MoSeS system is distinct from other systems in the following respects:

•MoSeS provides accurate, ad-hoc indoor positioning, thus no need for GPS
•MoSeS provides robust, ad-hoc networking among firefighters and commander, thus no need for 3/4G or WiFi
•MoSeS predicts health risks before they occur, so firefighters can be withdrawn before it actually happens

The MoSeS system is highly data driven. Not only sensors produce periodic data streams from firefighters to commander, also I/O data for display to firefighter heads-up display and text-to-speech flows between them. Many other data sources are foreseen in near future.

The data distribution between the applications on the smartphones of the firefighters and on the tablet of the commander is based on the Publish-subscribe paradigm because for its advantageous asynchronous properties. The project uses Vortex Cafe because they have Android smartphones and tablets in the system. MoSeS uses a MANET as the underlying wireless network. Though made robust, such ad-hoc wireless networks may have intermittent connectivity from time to time. To deal with this, our research focusses first on robust data distribution. In a second phase, we address security in MoSeS, where we want to differentiate access to data based on the type of data and the role of the sender/receiver.

Edge AI

Gaming

Healthcare

Railway

Robotics

Semiconductor Solution

Smart Manufacturing

AMD-based Solutions

Ampere-based Solutions

Arm-based Solutions

Intel-based Solutions

MediaTek-based Solutions

NVIDIA-based Solutions

NXP-based Solutions

Qualcomm-based Solutions

EdgeOpen™ Voices

Support

Downloads

Contact

Values

Company

Newsroom