The world’s top technology companies are racing to develop capabilities to allow computers to recognize photographs, control robots, understand spoken language and perform other tasks that until recently could only be accomplished by humans, according to an article in last week’s Wall Street Journal.
Beginning in the 1960s, scientists tried to accomplish with computers a human-like intelligence "from the bottom up." When achieving that kind of monolith proved challenging, the term "artificial intelligence" faded from view. AI didn’t go away, but its ambitions became more modest.
Apple’s Siri, for example, doesn’t claim to have the answer to everything. But it does a very good job of answering those type questions that humans most ask.
Today, improved capabilities and such things as so-called "deep learning" are setting off an AI-related arms race, according to a similar article in last week’s New York Times. Powerful companies such as Facebook, Google, IBM, Microsoft and Baidu have all trumpeted recent advances in vision and speech recognition.
Don’t think for a moment that this has nothing to do with the process industries. Whether labelled with the AI moniker or not, the joining together of these type capabilities with the Internet of things is having a powerful industrial impact that will continue for the indefinite future.
Most interesting is the range of social issues growing use of robots will bring to the workplace.
In related and other news:
Yokogawa, Shell next-gen platform
Yokogawa and Shell have jointly developed a platform for advanced control & estimation, a software suite that speeds up and simplifies the process of designing, deploying, and maintaining advanced process control applications.
The suite brings together advanced process control technology Shell acquired in operating its plants and the real-time control technology that Yokogawa has developed. Advanced control improves product yield and reduces energy consumption by maintaining process values such as temperature, flow rate, and pressure within a set range, but also keeping them as close as possible to their optimal set points.
This software suite brings the following functionality available within the same platform:
- multivariable model predictive control models the dynamic characteristics of plant responses and controlling multiple variables.
- soft sensing estimates quality in real time based on temperature, flow rate, pressure, and other values.
With dynamic characteristic response models, plant step response tests are typically executed over long periods of time to determine how device manipulation impacts process values. Now this process is automated so the optimum model is chosen expeditiously.
A hierarchical set of screens are provided as a standard feature. An advanced application includes modules for multivariable model predictive control, soft sensing, the interface with the production control system, calculation and other functions.
Rockwell, Cisco training
The convergence of operations technology with information technology is a big theme at Rockwell Automation the last few years, and its close working relationship with Cisco Systems insures it brings the Internet of things and the latest commercial advances, suitably modified, to a plant near you.
Last year, the companies jointly rolled out the Managing Industrial Networks with Cisco Networking Technologies (IMINS) training course and Cisco Industrial Networking Specialist certification. This first-of-its-kind course provided foundational skills needed to manage and administer networked, industrial control systems.
This year, Rockwell Automation and Cisco are unveiling the five-day, hands-on Managing Industrial Networks for Manufacturing with Cisco Technologies (IMINS2) course and CCNA Industrial certification exam. The course offers deeper analysis of EtherNet/IP architectures with industrial protocols, wireless and security technologies implementation, and advanced troubleshooting.
“The number of smart devices and objects that make up the Internet of Everything will double to about 50 billion in the next five years, according to Cisco studies," said Jeanne Beliveau-Dunn, vice president and general manager, Cisco. "Manufacturers who want to be at the forefront of this revolution must empower their employees today with the skillsets they’ll need.
Industrial IP Advantage, a coalition of companies including Rockwell Automation, Cisco and Panduit, also offers e-learning training to meet the emerging needs of control engineers and IT professionals tasked with deploying a secure, holistic network architecture.
Big Oil, Big Data
Ziebel AS, a provider of specialist well intervention services says it believes it’s made the largest-ever collection and transfer of data in the oil and gas industry. The collection, carried out on behalf of a major operator in the North Sea, was completed over an eight-month period. The volume of data gathered was 1,708 terabytes.
Advances in investigation and collection mean that oilfield data volume is expanding at a phenomenal rate. Industry experts believe that the sheer amount of data collected is growing by a factor of five every year. This results in new problems of data transfer. The data can’t simply be sent as an e-mail attachment or handed over on a USB stick. Instead, specialist equipment capable of temporarily storing and then transferring massive volumes of data to a client is essential.
During a recent operation in the North Sea, Ziebel collected data from a number of wells as part of an ongoing well monitoring regime. The data was collected using fiber optic composite rod technology to access and visualize wellbores in real-time. The data gathered was then compressed and transferred onto five servers, each of which comprises 36 hard drives. Each server measures 17.5 cm by 48 cm by 71 cm, and has a capacity of 120 terabytes. Four of the servers were filled to capacity, with the fifth approximately half-full. The data will be kept on these servers until the customer has uploaded it for analysis, whereupon it will be used to manage and develop wells in the future.
"What made this operation unique was its sheer size – I doubt that a bigger transfer has ever been carried out in the oil and gas industry," said Roger Sambrook, Ziebel. "They needed to collect essential data from a network of wells in the North Sea, and then store and transfer that data in an efficient, accessible way.
Founded in Stavanger, Norway in 2006, Ziebel provides specialist well intervention services to the global oil and gas industry.
Test-drive video walls
The Userful Network Video Wall control center is available online for anyone to access directly from their browser. Last December, Userful unveiled the world’s first video wall to deliver real time, 4k content over the network on up to twenty-five screens. Now the tool is online to publicly demonstrate how easy it can be to set up, deploy and manage artistic or grid video walls.
The Video Wall outputs a broad range of 4k content, including video, HTML5, 3D, Flash and more. The control center also ties in with a variety of third-party content management systems to enable real-time, and scheduled live content to the video wall. The can be accessed from a browser from any device, even a smartphone or tablet.
Userful has released a short instructional video showing how easy it can be to configure and manage a video wall, and in conjunction, has launched an online self-service test site. The control center makes it extremely simple to navigate, align and color-calibrate a video wall–either in a grid layout, or an artistic style video wall with a heterogeneous mix of individual displays placed at any angle.
The Video Wall connects up to twenty five displays to a single server over a standard Ethernet network via a small, but robust zero client device. The platform allows for full creative layout of displays, so customers can deploy a standard grid configuration or setup an mosaic artistic ‘art-wall’ layout with any angle rotation to grab any audiences’ attention.