DISCUSSION

The Medius Sensor Fusion System represents a third generation sensor fusion system. The system not only performs basic safety and security functions but is also  the only system capable of collision avoidance through advanced multi-sensor fusion. A technology that was formerly reserved for military electronics is now available to provide simple, easy to comprehend situational awareness for today’s driving public.

Our Multi-Sensor Fusion technology combines measurements from a wide variety of sensors in much the same fashion as humans work with their senses. Our patented technology accesses possible objects as raw measurements from both on-and off-board sensors. We combine those measurements using a multiple-hypothesis adaptive tracking strategy. Our engineering staff and technology partners have over many years of development experience in this technology, tracking objects far less cooperative than the numerous vehicles and pedestrians, motor cycles, etc. that operate on today’s roads and highways.
Medius' system performs both cooperative and non-cooperative object tracking and ID. We do this in two ways, one is through the exploitation of multiple sensors in our vehicle, and the second is by sharing our kinematic data with other vehicles similarly equipped. This advanced sharing of information is derived from the military sector and requires no additional hardware from today’s stand-alone systems. Where other systems experience RF and IR coordination issues, we utilize those challenging conditions to improve our tracking solution.

The Medius approach has been to provide a system-of-systems integration, which provides full 360 ° monitoring around the vehicle. This approach is accomplished in a distributed system that utilizes generic sensor interfaces. Use of this generic plug-and-play approach allows sensors to be upgraded, added, or even removed easily while maintaining our hardware and software architectures.

This approach allows the system to migrate to a series of software modules that can easily be added to increase functionality as time goes on without frequent upgrades to the system hardware. This proven approach is not revolutionary, but rather has been common in the military sector and is flying in AWACS, U.S. and European fighters today. The sensing system operates in its environment in the same way humans process data from their senses. We utilize a Multiple Hypothesis Tracking (MHT) theory that adapts its kinematic state estimation based on maneuvers of the reported target. This means that targets from animals to humans to cars can be tracked accurately and simultaneously. This technology has been in development for years and has been optimized to track hundreds of targets with comparatively low CPU and memory utilization in a single module.

Through the use of our tracking technology, we can actually reduce the number of required sensors and use sensors of lesser performance because we task the system and give sensors cues about where to look to speed up processing and to maximize front end timeline utilization. Our system has key decision points that cause it to prioritize objects. These decision points represent time boundaries within which the system must perform in order to give the driver time enough to respond. 

This approach allows the system to respond with predetermined reaction times; or it can respond to the user over a period of time by measuring user reaction times through his/her interactions with the vehicle (i.e. braking, turning, response to cut-ins, etc.)

As the vehicle moves down the road, it obtains information, from on-board sensors and systems, about objects in its environment. It then passes the data to its Human Machine Interface for updating the situation awareness picture. This situation awareness picture is then presented to the operator in a simple, intuitive user interface consisting of video, haptic and/or audio cueing, depending on the functions enabled.

Use of multiple sensors gives the driver the ability to deal with extreme and often dangerous driving conditions including rain, snow, extreme fog conditions, night or day. These capabilities act as extensions of the driver’s senses and help improve reaction times.
 

SEE FURTHER DISCUSSION AT:  INITIATIVES

PATENTS | DOWNLOADS | NEWS & EVENTS | STATS

COPYRIGHT © 2007 MEDIUS INC, SEATTLE, WA