The Big Semiconductor Water Problem
Global Lighthouse Network: Unlocking Sustainability through Fourth Industrial Revolution Technologies
The Global Lighthouse Network is a community of production sites and other facilities that are world leaders in the adoption and integration of the cutting-edge technologies of the Fourth Industrial Revolution (4IR). Lighthouses apply 4IR technologies such as artificial intelligence, 3D-printing and big data analytics to maximize efficiency and competitiveness at scale, transform business models and drive economic growth, while augmenting the workforce, protecting the environment and contributing to a learning journey for all-sized manufacturers across all geographies and industries.
Why Robots Can’t Sew Your T-Shirt
But sewing has been notoriously difficult to automate, because textiles bunch and stretch as they’re worked with. Human hands are adept at keeping fabric organized as it passes through a sewing machine. Robots typically are not deft enough to handle the task.
SoftWear’s robots overcame those hurdles. They can make a T-shirt. But making them as cheaply as human workers do in places like China or Guatemala, where workers earn a fraction of what they might make in the US, will be a challenge, says Sheng Lu, a professor of fashion and apparel studies at the University of Delaware.
SoftWear calls its robotic systems Sewbots. They are basically elaborate work tables that pair sewing machines with complex sensors. The company zealously guards the details of how they work, but here are the basics: Fabric is cut into pieces that will become parts of the shirt: the front, the back, and the sleeves. Those pieces are loaded into a work line where, instead of a person pushing the fabric through a sewing machine, a complicated vacuum system stretches and moves the material. Cameras track the threads in each panel, allowing the system to make adjustments while the garment is being constructed.
How DeepMind is Reinventing the Robot
To train a robot, though, such huge data sets are unavailable. “This is a problem,” notes Hadsell. You can simulate thousands of games of Go in a few minutes, run in parallel on hundreds of CPUs. But if it takes 3 seconds for a robot to pick up a cup, then you can only do it 20 times per minute per robot. What’s more, if your image-recognition system gets the first million images wrong, it might not matter much. But if your bipedal robot falls over the first 1,000 times it tries to walk, then you’ll have a badly dented robot, if not worse.
There are more profound problems. The one that Hadsell is most interested in is that of catastrophic forgetting: When an AI learns a new task, it has an unfortunate tendency to forget all the old ones. “One of our classic examples was training an agent to play Pong,” says Hadsell. You could get it playing so that it would win every game against the computer 20 to zero, she says; but if you perturb the weights just a little bit, such as by training it on Breakout or Pac-Man, “then the performance will—boop!—go off a cliff.” Suddenly it will lose 20 to zero every time.
There are ways around the problem. An obvious one is to simply silo off each skill. Train your neural network on one task, save its network’s weights to its data storage, then train it on a new task, saving those weights elsewhere. Then the system need only recognize the type of challenge at the outset and apply the proper set of weights.
But that strategy is limited. For one thing, it’s not scalable. If you want to build a robot capable of accomplishing many tasks in a broad range of environments, you’d have to train it on every single one of them. And if the environment is unstructured, you won’t even know ahead of time what some of those tasks will be. Another problem is that this strategy doesn’t let the robot transfer the skills that it acquired solving task A over to task B. Such an ability to transfer knowledge is an important hallmark of human learning.
Hadsell’s preferred approach is something called “elastic weight consolidation.” The gist is that, after learning a task, a neural network will assess which of the synapselike connections between the neuronlike nodes are the most important to that task, and it will partially freeze their weights.
What is Indirect Material Optimization, and How Can It Turbocharge Your Supply Chain?
The problem of poor indirect material management stems partly from the inability of legacy and manual data management systems to quantify and measure indirect materials. These systems focus on the movement of trackable, direct materials but often ignore the indirect materials that are not utilized in the finished product.
This problem is further compounded by the fact that some organizations may not have internal controls in place to categorize and attach value to indirect spending. There might also be a lack of employee education about how and why indirect materials must be managed.
Industry analysts estimate that US manufacturers could save over $5 trillion annually just by optimizing their indirect material management processes.
Hybrid machine learning-enabled adaptive welding speed control
This research presents a preliminary study on developing appropriate Machine Learning (ML) techniques for real-time welding quality prediction and adaptive welding speed adjustment for GTAW welding at a constant current. In order to collect the data needed to train the hybrid ML models, two cameras are applied to monitor the welding process, with one camera (available in practical robotic welding) recording the top-side weld pool dynamics and a second camera (unavailable in practical robotic welding, but applicable for training purpose) recording the back-side bead formation. Given these two data sets, correlations can be discovered through a convolutional neural network (CNN) that is good at image characterization. With the CNN, top-side weld pool images can be analyzed to predict the back-side bead width during active welding control.