In the world of science fiction, it is an old, cautionary story. An autonomous, intelligent robotic defense system goes insane — or simply misinterprets its instructions — and attacks everyone, including the civilization it was designed to defend.

For defense planners today, this scenario is no longer pure science fiction. No, the Pentagon is not planning giant orbital robot battle stations that might erroneously attack Earth — but artificial intelligence (AI) technology, and especially “deep learning,” are increasingly being applied to real-world defense tasks such as analyzing reconnaissance images.

In order to rely on information detected by such sophisticated technology, defense planners and commanders need to understand how deep learning works. And precisely because it is such a powerful technology, fully understanding it turns out to be a very challenging task.

Deep Learning and Image Recognition

The technology of deep learning, as Will Knight reports at MIT Technology Review, is inspired by the complex architecture of the human brain. The AI software simulates the multiple layers of neurons and synapses, and also simulates the learning process of the human mind.

A classic example of the deep-learning process is teaching the system to recognize a cat. The system is provided with a large library of imagery, some of which includes images of cats. In a “tuning” or training process, the system attempts to identify whether or not there is a cat in an image, and records a positive score whenever its human trainers confirm that there is in fact a cat in the image.

Because the system can swiftly crunch through vast numbers of training images, it can soon learn to identify cats with high reliability.

Identifying cats is rarely a vital security issue. But the same technology can be used for other image recognition tasks — for example, to scan thousands or millions of satellite images for evidence of tanks or missile launchers. Hence, the interest of defense planners in deep-learning technology.

How Do You Recognize a Cat?

But this is where things get challenging. While deep-learning technology can be taught to recognize a cat — or a missile launcher — it cannot tell us how it performs this accomplishment. Says Knight, “it isn’t clear whether the system may be focusing on the whiskers, the ears, or even the cat’s blanket in an image.”

Sometimes this is not a problem. Perhaps the AI system can pass its results on to human intelligence analysts, who can double check to make sure the AI is correct in its identification of, say, missile launchers. But in an urgent crisis situation, there might be no time to wait for direct human confirmation — the launchers need to be identified and engaged now.

In that situation, planners and commanders need to know exactly how much confidence they can place in an AI system. Which means learning how the AI learns, and understanding its (simulated) “thought” processes.

Understanding how deep learning works is so challenging that the Defense Advanced Research Projects Agency (DARPA) has no fewer than 13 different research projects underway, using a range of different techniques to improve our understanding of the deep-learning process.

Which also explains why Northrop Grumman — a traditional leader in drones, autonomous systems and defense robotics — is at the forefront of research into understanding deep learning.

The post Deep Learning, and the Art of Understanding Robots appeared first on Artificial Intelligence.

Technavio’s latest market research report on the deep learning market in the US provides an analysis of the most important trends expected to impact the market outlook from 2017-2021. Technavio defines an emerging trend as a factor that has the potential to significantly impact the market and contribute to its growth or decline.

According to Bharath Kanniappan, a lead analyst at Technavio for robotics research, “The deep learning market in the US is expected to grow at a phenomenal CAGR of over 57% during the forecast period. Industries across the US are striving hard to channelize and optimize multiple facets of operations, including data analysis, storage, strategy, and decision-making. Deep learning has surfaced as a powerful tool that assists industries in improving the programming of automated machines/equipment and inducing self-learning capabilities.”

The top three emerging market trends driving the global deep learning market in the US market according to Technavio research analysts are:

• Advances in deep learning
• Reinforcement learning
• Combating security threats using AI

Looking for more information on this market? Request a free sample report

Technavio’s sample reports are free of charge and contain multiple sections of the report including the market size and forecast, drivers, challenges, trends, and more.

Advances in deep learning

With industries harnessing deep learning technology to optimize operations and make real-time decisions, modular capabilities in deep learning will aid visual design, configuration, and training new models obtained from pre-existent building blocks. A major structural change will emerge as a result, known as transfer learning, which will enable experiential solving of similar cases.

As deep learning technology gets adopted by masses, the market will progress toward a self-service cloud-enabled delivery model. This cloud-based platform will deliver fast results and would be useful in overcoming technical difficulties encountered in deep learning algorithm. This evolution in deep learning market will pave the way for a new wave of industrial revolution. Industries will switch from their traditional mode of disconnected systems and reactive approach to an integrated and proactive approach based decision-making.

Reinforcement learning

Reinforcement learning is a specialized form of supervised learning with a provision of training information provided by the environment. The learner/user in reinforcement learning needs to uncover actions that generate the best results, by being a part of the decision-making process. Instead of following instructions, the learner can override the system-generated commands to take decision on its own. Reinforcement learning is an evolved version of machine learning and superior in terms of results delivered. Unlike supervised learning, reinforcement learning exhibits adequacy in situations when there is an absence of a knowledgeable supervisor. In such unfamiliar situations, an agent is required to be able to learn from the interface and by using its own experience. This is where reinforcement learning is expected to showcase its advantages.

Combating security threats using AI

Leakage of sensitive information and security threats are some of the major problems faced by end-users while deploying automation solutions. In recent times, several instances of cyber security concerns were reported in manufacturing industries such as oil and gas, automotive, pulp and paper, chemical and petrochemical, food and beverages, and pharmaceutical.

Traditional systems that ensure cyber security ar reliant on signature-based detection, network perimeter security model, and firewalls. However, this approach is not a highly robust method as continuous exchange of confidential information taking place through emails and websites is potentially vulnerable to malware.

“AI technologies can help end-users address several issues related to cyber-attacks including firewall failure, security threat to voluminous sensitive data, and scalability challenges. Advanced security products that are based on AI-technologies like deep learning can recognize and destroy malware rapidly during its development. In this way, businesses can ensure security and integrity of their critical data,” says Bharath.

Browse Related Reports:

• Global Laboratory Plate Handling Systems Market 2017-2021
• Global Robotic Pool Cleaner Market 2017-2021
• Global Fiberglass Cutting Robots Market 2017-2021

About Technavio

Technavio is a leading global technology research and advisory company. Their research and analysis focuses on emerging market trends and provides actionable insights to help businesses identify market opportunities and develop effective strategies to optimize their market positions.

With over 500 specialized analysts, Technavio’s report library consists of more than 10,000 reports and counting, covering 800 technologies, spanning across 50 countries. Their client base consists of enterprises of all sizes, including more than 100 Fortune 500 companies. This growing client base relies on Technavio’s comprehensive coverage, extensive research, and actionable market insights to identify opportunities in existing and potential markets and assess their competitive positions within changing market scenarios.

The post Deep Learning Market to Help in Combating Security Threats appeared first on Artificial Intelligence.

Through the Looking Glass

When we were in the city of Danyang, China, we witnessed a real-life paradox. Danyang is best known for its explosive growth in optical lens manufacturing over the last decade, sprouting hundreds of factories with cleanrooms chock-full of gleaming, automated machinery. There is a lot that goes into manufacturing a lens, and this machinery performs the bulk of it: from lens curing, lens cleaning, to lens coating. As Stanford AI researchers and engineers wildly interested in Chinese manufacturing, we were impressed by this caliber of automation.

But the paradox greeted us as soon as we walked into the rooms responsible for the most critical part of the entire process: quality control. Both before and after lenses enter a series of complex machines, they must be inspected by throngs of female workers whose sole job is to sit at a bench and perform the following:

1. Grab a rack of lenses from a mound of racks piled in the center of the room

2. Grasp each lens by the rim and tilt it against a backlight, while checking for digs, dark spots, scratches, or air bubbles

3. Place passed lenses back into the rack, and move the racks into a staging area for the next step

4. Repeat

In each of the subsequent optics factories we visited, this process presented itself over and over. Naturally, we asked the factory managers about this process, and received a slew of complaints. We condensed them into five main points:

1. A worker takes five seconds to inspect a single lens, which is too slow.

2. Different lenses have different lens properties and thicknesses, and factory workers must constantly adjust to the change, leading to inaccuracy.

3. Quality control workers get tired of staring at lenses all day and accidentally allow more flawed lenses to pass as the day goes by.

4. Machines to ascertain lens light properties are in wide usage but no machines exist to inspect spectacle lenses for surface flaws and defects. This is a need.

5. Once a flawed lens is introduced into a batch, the entire batch must be inspected all over again to remove flawed lenses. This dramatically increases cost.

One factory quality control manager claimed that over 40% of customer complaints were due to errors in lens cosmetic defects. Others claimed that the optical lens quality control process is a holy grail for automation because it is so endemic to the entire optics industry.

A Broader Trend

Quality control was endemic to almost all the factories we visited in other verticals. Beyond optics, we visited countless plastics manufacturers, medical device manufacturers, and automotive parts suppliers all facing very similar problems. Many of these factories automated everything in sight, from injection molding to assembly, with armies of Kuka and Fanuc robotic arms. But quality control continues to present an enormous challenge due to its reliance on human-level visual understanding and adaptation to constantly changing conditions and products. Many factories resort to hiring workers to perform this crucial step.

But human labor in quality control is eroding. A constant concern among factory owners is how difficult it is to hire new workers due to decreased labor supply and the ever-increasing cost of labor in China. According to the U.S. Bureau of Labor Statistics, hourly compensation costs in manufacturing in China nearly tripled from 2002-2009. In fact, one of the biggest types of questions we got from factory bosses was “how can we automate this process?” Where this was a task with high variability, ripe for the application of artificial intelligence.

China is determined to answer this question. In its most recent Five Year Plan, the Communist Party of China firmly encouraged the union of artificial intelligence and manufacturing, and deemed China’s manufacturing future as “Made in China 2025,” where autonomous manufacturing or manufacturing of incredibly mission critical systems would take center stage. Such support would mean pouring billions of dollars into AI research, as well as providing subsidies for factories that embed AI into their operations.

Tackling Manufacturing Problems with Intelligence

To automate processes that require human level intelligence, we require algorithms and technologies with human-level intelligence. Thankfully, researchers from top institutions and companies have developed a myriad of techniques that can dramatically improve automation.

One such technology that falls under artificial intelligence is computer vision aided by convolutional neural networks (CNNs). CNNs can automatically learn what distinguishes good parts from not good parts on an assembly line with incredible speed. With a good AI product dedicated to quality control, and training images that depict good parts and not good parts, training a CNN can be quite fast. This is an incredible solution for high-mix environments where product environments are constantly changing, and time is valuable. In optics especially, CNNs can quickly respond to varying lens properties and prevent errors and downtime.

In low-mix environments, traditional computer vision technologies can make an enormous difference as well by performing a single task very well and very quickly, albeit with less flexibility. Furthermore, advances in reinforcement learning are also teaching robots to grasp objects properly, and even learn from human actions in simple tasks. This lends itself towards pick and place or general purpose robots of the future that can learn on their own, without having to be constantly reprogrammed for each new product requirement. In optics manufacturing, robots can be trained to perform the difficult task of filling lens molds with polymer, which currently requires experienced and trained workers. We will elaborate on these technologies in a later series of posts.

Valkri Is Tackling This Problem in Optics with AI

After coming back from Danyang, we explored this problem further. Weeks of testing and development revealed that a machine vision system with dark field lighting, with images analyzed by deep learning technology, can unlock the promise of efficient and strikingly accurate quality control in optics manufacturing. We are partnering with Tsinghua University to realize this technology and bring it to reality to optics factories worldwide.

With a solution like ours, optics factories can automatically inspect lenses that either arrive on an assembly line or are placed into the system by a quality control worker. They can ensure that both external and internal defects are detected with incredible precision and speed, with minimal human interference or error. This can allow production speed to escape the limitations imposed by the human eye, and enable a smooth supply chain on the brand and distributor side.

The post Using Artificial Intelligence to Improve Quality Control appeared first on Artificial Intelligence.

Keeping pace with trends in any industry is an integral point for loyalty marketers.

When it comes to loyalty program science, Loyalty360 talked to Eoin O’Sullivan, global director of analytics/bBI at Snipp, to find out the latest trends that loyalty marketers should be aware of and incorporate into their respective business strategies.

Can you talk about some of the new trends in loyalty program data science?
O’Sullivan: Two big new trends are the integration of data science into application and the democratization or demystification of data science. Traditionally, data science was a standalone process. The data scientist would receive data in a file, work their magic, return with some predictive analytics, and then the company would make strategic decisions based on this. One of the big new trends is to build and embed this type of analytics into the loyalty program applications and processes. Once you productionalize this type of predictive analytics, you can then move on the final frontier of analytic capabilities, prescriptive analytics, this means that your Loyalty application can make suggestions to take advantage of your stored models and analysis.

A lot of companies such as Tableau, Microsoft, and Oracle are embedding data science capabilities into existing products. This is helping bring the data science toolkit into more easily maintainable and production ready platforms, increasing the pervasiveness of data science. Another benefit of these enhanced traditional tools is they remove a lot of the barriers to entry into the realm of data science, enabling a new breed of analysts called citizen data scientists.

How can these trends help loyalty marketers today?
O’Sullivan: There are numerous ways. Loyalty program applications with embedded analytics systems could, for example, automatically add a customer to a custom segment based on demographic information (age range, gender, zip) or move a customer between custom segments based on historical activity (e.g. high-value infrequent transactions). This will allow loyalty marketers to automatically offer more nuanced and focused incentives to these segments. Rather than a one-size-fits all offering, out of the box, you would have a maintained customer segment that will respond best to promotions, a segment that responds best to sweepstakes and so on.

Embedded analytics systems could also provide a real-time recommendation engine based on market basket analysis (association rules) or churn alerts, which would notify clients about any potential flight risks within the loyalty program and take steps to increase the chances of retention.

The increasing pervasiveness of data science will allow loyalty businesses to become more data-enabled. I don’t like the term “data-driven” as this takes away the crucial “common sense” element that an experienced marketer provides. When you are data-enabled, you can prioritize and incorporate findings from data science into your campaigns increase the change of success.

The democratization of data science will allow the more technically abled loyalty marketers to take advantage of the more sophisticated tools without having to have a programming background.

They can then use these tools to assist in answering “the why”, giving the client more information and an insight into as to why certain campaigns are succeeding others are not. This increased transparency will help the marketer guide the client in a direction that will increase chances of future programs succeeding.

What is being done well in loyalty program data science and where do the challenges lie?
O’Sullivan: There is a nicely defined and establish set of algorithms for the Loyalty industry for use in churn, recommendation, and segmentation. So, once you have your data in a good place, and you have the appropriate technology in place, then the barriers to creating smart campaigns are nowhere near where they used to be.
The challenges lie in the quality and completeness of this data, as the volume and variety of data increase in this age of big data the quality controls around this data is not always of the highest standard. You can’t just drop an expensive piece of software on just any data and expect results. You must take the process of data cleaning and data quality seriously. You also need to model, integrate, and augment data from a huge variety of sources such as web analytics, social media, demographic data, weather data in addition to the data from your core systems.

There is also the issue of being data-driven vs data-enabled. A good analogy that I have seen is to flying an airplane. As good as the automation technology is when critical decisions must be made it is a highly trained pilot that you want making those decisions. Data science, no matter how pervasive it becomes, should not be an excuse for abdicating decision making. It should help you make better decisions.

Why is loyalty program data science so important?
O’Sullivan: It is so important to understand what is happening or what has happened via historical reporting and this traditional BI/reporting requirement will never cease to exist. To take loyalty programs to the next level and create that sales lift that marketers need to have, however, then you need to step beyond traditional operational reporting and into the area of predictive and prescriptive reporting.

It’s a long-held truism that retaining customers is vastly more efficient than gaining new customers and you do this by building out quality Loyalty programs. To achieve and maintain quality loyalty programs, you will need to understand the why behind loyalty program success or failure. Using data science to find the reasons behind a successful or unsuccessful program, and using data science to greatly enhance the chances of a successful or unsuccessful program, is a hugely important means of building a strong and lasting relationship with your customers.

What do you foresee in the future for loyalty program data science?
O’Sullivan: The technology for the future of loyalty data science is in place right now, however, in a similar fashion to the way automation in the motoring industry is gradually becoming mainstream, the same will happen with loyalty programs.

We are not too far away from a place where loyalty program data science technology will work in a symbiotic relationship with the marketer. Marketers will have their productivity enhanced with automation based on data science. Deep learning technology will continue to enable big leaps forward in loyalty program data science. This will allow customers to interact with a program’s embedded chatbots in very natural fashion and will enable advanced sentiment analysis, facial recognition, and image classification, allowing marketers get age and gender information from person’s picture in a similar fashion to Microsoft’s

As this intelligence becomes more embedded in our applications and we more move to near real time predictive analytics, we will have modules that will prompt loyalty marketers to change incentives and distribution medium while the campaign is in flight. For instance, if there is a shift in the demographics of the customer base in a loyalty program, then the marketer could be prompted to change the communication medium to a different social media channel to obtain a 5 percent increase in participation.
Interesting and exciting times ahead for all in this industry.

The post New Trends in Loyalty Program Data Science appeared first on Artificial Intelligence.