In 1913, Henry Ford revolutionized automobile production with the first mobile assembly line, an innovation that made building a new model faster and more efficient. Hundreds of years later, Ford is using artificial intelligence to speed up today̵7;s production lines.
At Ford’s power plant in Livonia, Michigan, the station where robots assemble the torque converter now has an AI-powered system to learn from previous attempts to effectively wiggle parts into place. Top performance Inside a large safety cage, the robot arm wheels grab circular metal pieces, each about the diameter of a dinner plate from the conveyor belt, and plug them together.
Ford used a technology from a startup called Symbio Robotics, which looked at the past few hundred efforts to determine which approaches and moves appeared to work best. A computer sitting out of a cage showing Symbio’s technology for detecting and controlling the arm Toyota and Nissan use the same technology to improve production line efficiency.
The technology allows this part of the assembly line to run 15 percent faster, a significant improvement in automotive production where thin margins are highly dependent on productivity.
“Personally, I think it will be something in the future,” said Lon Van Geloven, production manager at the Livonia plant. He said Ford plans to explore whether to bring the technology to other plants. Van Geloven says the technology can be used anywhere, so computers learn from a feeling of how things fit together. “There are a lot of these applications,” he said.
AI is often viewed as a disruptive and transformative technology. But Livonia’s torque settings show how AI may creep into industrial processes in gradual and often invisible ways.
Automotive manufacturing is already an automated system. But robots that help in assembling, welding and painting vehicles are powerful and precise robots that repeat the same work endlessly. But lack the ability to understand or respond to the environment
Adding automation is challenging. Jobs that have yet to reach the machines include tasks such as flexible feeding of wires through the car’s instrument panel and body.In 2018, Elon Musk blamed the Tesla Model 3’s production delay over the decision to rely on the system. More automation in production
Researchers and startups are exploring ways AI will empower robots, such as enabling them to recognize and capture even unfamiliar objects that move along a conveyor belt. Ford’s example shows how existing machines can often be improved by introducing simple detection and learning capabilities.
“This is very valuable,” said Cheryl Xu, a professor at North Carolina State University who works in manufacturing technology. She added that her students are exploring ways machine learning can improve the efficiency of automation.
One of the major challenges, Xu said, is that each manufacturing process is unique and requires a specific type of automation. Some machine learning methods may be unpredictable, she noted, and the increasing use of AI poses new cyber security challenges.
Timothy Chan, a professor of mechanical and industrial engineering at the University of Toronto, said the potential of AI to customize industrial processes is huge. Computer vision algorithms can be trained to detect defects in products or problems in production lines. Similar technologies can help enforce safety rules by identifying when someone is not wearing the correct safety equipment, for example.
Chan said a major challenge for manufacturers is integrating new technology into workflows without disrupting production efficiency. He also said it can be difficult if employees are unfamiliar with working with advanced computer systems.
This doesn’t seem like a problem in Livonia, Ford’s production manager Van Geloven believes consumer devices such as smartphones and game consoles make workers more technologically savvy. And for all the talk about AI taking on blue-collar jobs, he noted that this is not a problem when using AI to improve the efficiency of existing automation. “Manpower is really important,” he said.
This story originally appeared on the WebWire.com.