An issue I haven’t talked much about is the economic and
technical feasibility of the 4-Dimensional Global Maps, and other services, I’ve
been discussing.
I was watching a YouTube video of the Intelligent
Transportation System conference
in Vienna Austria in 2012, including a black-tie ball, and looking at the
website of the Car2Car
consortium of European auto manufacturers and suppliers. That led me to focus
on the amounts of money being spent by the auto manufacturers,
suppliers, governments, and others, including Google, to design and implement these
systems for autonomous cars.
Thus figuring out how to fund the development and
implementation of the algorithms and code, as well as the processing, storage, and operation is essential for the continued viability of my ideas.
So who would spend the money to develop the systems I’ve
been discussing? Fortunately, 4-Dimensional Global Maps are already in
widespread use in MMOGs
(Massively Multi-player Online Games). World of Warcraft,
has about 9M subscribers, demonstrating the feasibility and viability of very large systems.
The MMOGs have already developed systems for creating and
managing 4-dimension maps, including scenery and motion of many objects simultaneously,
spread among large numbers of separate processors. Extensive competition among
games has optimized communication, shared storage and processing, and features
such as zooming in and out. They use physics models to
predict motion so that you don't have to send updates unless there is a
significant change in the motion of the objects (i.e., a significant change in
acceleration, in 2 or 3 dimensions).
These models provide detailed local maps maintaining up-to-date characteristics
of the local area, while also providing aggregation of information for larger
scale maps in the distance. Thus the algorithms and even the code are already
available, with several platforms available,
and more under development.
A next question is who will pay for the processing and
storage for these maps. We can't count on having essentially free unlimited
processing and storage from the likes of Google for our future system. So
having the processors which are running the autonomous systems also provide the
services in a distributed manner establishes feasibility and economy for the
whole system.
“Volunteer Computing” has been designed to run applications
on myriad distributed PCs over the Internet, providing massive processing and storage,
while maintaining privacy and security. A major example is SETI@home, with 228K active
hosts and 3M total hosts worldwide, producing an average of 505 TFLOPS.
SETI@home uses the BOINC open
source middleware system, thus demonstrating the potential for both distributed
processing and open source coding.
Widely used examples of open source coding
are LINIX, Java, MySQL, Apache web server, and Firefox. These paradigms establish the feasibility of the overall development of our Autonomous systems.
Thus each Autonomous Processor will participate in
providing the overall services of computing, verifying, storing and processing
maps and routing information.
We also anticipate including personal communication devices,
smartphones, digital assistants, etc. as participants in the monitoring and
mapping system. They add not only processing and storage to the system, but
provide additional Points-of-View and sensors, thus filling out the needed
information for accurate and timely 4-Dimension Global Maps and other services,
even if no Autonomous Vehicles are in the area.
No comments:
Post a Comment