Motionloft to Power Real-Time Data Acquisition in Cities Using NVIDIA Jetson
Apr 03, 2017, 09:53 ET
SAN FRANCISCO, April 3, 2017 /CNW/ -- Motionloft is providing businesses and cities with detailed, real-time data to help cities get deeper insights into stationary and moving objects for indoor and outdoor environments.
Motionloft's sensor includes an NVIDIA Jetson TX1 supercomputer on a module running advanced neural networks, along with an HD stereo camera and a range of wireless radios. This sensor is capable of near-instantaneous scene analysis, transmission of resulting metadata and real-time critical event notifications.
Using NVIDIA's technology, Motionloft ensures customer privacy because video data streaming never happens outside the sensor itself. Video is analyzed on-site and then is immediately removed from the embedded system. The resulting metadata bitstream describes the scene but cannot be used to reconstruct it.
The Jetson TX1's high performance and power efficiency, paired with Motionloft's wide range of proprietary algorithms, ensures a high level of accuracy and flexibility. Motionloft sensors can be updated remotely to perform a variety of tasks such as estimating traffic patterns and performing detailed pedestrian and object counts. This includes their precise trajectory, regardless of whether that person or object is stationary, moving fast or slow. These sensors can also act as self-contained analysis units or their powers can be combined to create unlimited fields of view.
"The Jetson TX1 is the best technology on the market for our product, providing Motionloft with the ability to analyze complex data in real time for highly accurate behavioral and traffic data insights," said Joyce Reitman, CEO of Motionloft.
Motionloft can provide invaluable insights, including:
- Customer dwell time of how long a person spends in front of an object and what they did next.
- Customer capture rate of how many people decided to enter a store, was it before or after seeing something of interest in a shopping window.
- Customer conversion rate of how many captured customers went through the checkout, length of the checkout queue line, time it took to complete the purchase.
- Traffic queue counts, time spent in front of toll booths and traffic lights, vehicle speed and trajectory, data points in understanding outdoor vehicle traffic.
- Pedestrian and vehicular interactions at intersections to address safety improvements.
- Number and location of parking spaces available in multi- or single-story structures, necessary for optimizing parking utilization.
- Number of people waiting at bus stops along with entering and exiting them for a true smart city application.
"Jetson is the best platform for AI at the edge and Motionloft is using it to provide advanced real-time analytics with computer vision-based sensor technology," said Deepu Talla, vice president and general manager of the Tegra business at NVIDIA. "Leveraging Jetson, Motionloft is using deep learning to help its customers gain more visibility into pedestrian and vehicle traffic to make better informed decisions."
Motionloft at ISC West 2017
Motionloft will be inside NVIDIA's booth #20075 at ISC West in Las Vegas, from April 5-7. The company will showcase its complete solution of software, vision algorithms and hardware utilizing NVIDIA Jetson. Christopher Garrison, Vice President of Business Development, will be doing an in-booth theater talk on April 6 from 1:00-1:15pm PT on the topic of "Understanding Real World Activity Through Computer Vision."
Motionloft is an investment of Mark Cuban Companies and is based in San Francisco, California. Motionloft provides data analytics serving pedestrian and vehicle real time counts to many of the largest cities, retailers and property owners in the world. Motionloft is disrupting how physical counting is done in the real world. For more information visit http://www.motionloft.com.
For further information: Joyce Reitman - CEO, (415) 625-0162, http://motionloft.com
Share this article