Follow Datanami:
August 16, 2016

Ford Accelerates Driverless Car Effort With Machine Learning

A key component driving the development of driverless cars is machine learning and other artificial intelligence capabilities along with computer vision approaches used for image and signal processing. Ford Motor Co., which is targeting fully autonomous vehicles for ride sharing by 2021, unveiled a series of machine learning and machine vision deals as it doubles the size of it Silicon Valley research campus.

The U.S. carmaker (NYSE: F) announced an acquisition and others investments on Tuesday (Aug. 16), including a deal to buy SAIPS, an Israeli computer vision and machine learning company. Ford also disclosed a licensing deal with machine vision specialist Nirenberg Neuroscience, who is credited with cracking the code the eye uses to transmit visual information to the brain.

Furthering its autonomous vehicle initiative, Ford also announced an investment in the 3-D mapping startup, Civil Maps. The company said the moves would double the size of its Palo Alto research team by the end of 2017.

The automaker’s autonomous vehicle push is part of an overarching strategy to become a leader in mobile connectivity along with data and analytics. Ford’s industry standard driverless vehicles for ride sharing and “hailing” would eliminate the steering wheel along with gas and brake pedals. It is currently testing self-driving Fusion Hybrids in Arizona, California and Michigan. It expects to triple the number of self-driving cars being tested next year.

Founded in 2012, Tel Aviv-based SAIPS is a machine-learning specialist that has developed algorithmic engines based on deep learning that would be used for image and video processing. Ford said the technology would help autonomous vehicles learn and adapt to their surroundings.

Ford also said it has licensed Nirenberg Neurosciences machine vision platform used for navigation, object recognition and other applications. “Ford’s partnership with Nirenberg Neuroscience will help bring humanlike intelligence to the machine learning modules of its autonomous vehicle virtual driver system,” the partners said.

The carmaker also invested in Civil Maps to further development of high-resolution, 3-D mapping capabilities. Civil Maps touts a new 3-D mapping technique that it claims is scalable and more efficient than existing processes. Ford said the mapping capability would provide it with an alternative approach for developing 3-D maps of autonomous vehicle routes.

“Ford will be mass producing vehicles with full autonomy within five years,” predicted Ford CEO and President Mark Fields. “We see the autonomous car changing the way the world moves.”

The carmaker added that it is working with more than 40 startups to develop autonomous vehicle applications and services. It opened its Palo Alto research center in January 2015.

Underpinning efforts by Ford, Google (NASDAQ: GOOGL) and others to mass-produce self-driving cars is the massive streams of position data that must be extracted and loaded via techniques such as event stream processing used to automate car maneuvers.

Recent items:

Five Answers for Finding Bigger Insight from Your [Big] Data

How Nvidia is Unlocking the Potential of GPU-Powered Deep Learning