by Gary Mintchell | Apr 24, 2024 | Generative AI
A personal development speaker I often listen to delivers a set of talks on developing personal guardrails designed to prevent us from going off the deep end emotionally and relationally. Similarly as we explore this new age of artificial intelligence (AI) people are recognizing that we could use a set of guardrails to help guide our collective decisions using this new technology.
Collective guardrails generally include social norms, laws, and rules. Do we have any existing guardrails that will help us navigate AI? Where might they come from? What guardrails might work? Which might fall short?
Guardrails: Guiding Human Decisions in the Age of AI by Urs Gasser and Viktor Mayer-Schönberger came out recently. I promised to read and review it a couple of months ago. It got buried amongst other reading, plus it is not one of those “skim through” business books. This book has real meat. Based on the latest insights from the cognitive sciences, economics, and public policy, Guardrails offers a novel approach to shaping decisions by embracing human agency in its social context.
The authors with meticulous research lead us through technology approaches and social approaches through laws and regulations revealing the benefits but also the shortcomings of each.
From the press release: In this visionary book, Urs Gasser and Viktor Mayer-Schönberger show how the quick embrace of technological solutions can lead to results we don’t always want and explain how society itself can provide guardrails more suited to the digital age, ones that empower individual choice while accounting for the social good, encourage flexibility in the face of changing circumstances, and ultimately help us to make better decisions as we tackle the most daunting problems of our times, such as global injustice and climate change.
They conclude, “We hope that our readers—and everyone in governments, companies, and communities tasked with confronting some of humanity’s biggest challenges—will embrace this timely opportunity to think about and experiment with smarter guardrails to work toward better, fairer, and more sustainable futures.”
Urs Gasser is professor of public policy, governance, and innovative technology and dean of the School of Social Sciences and Technology at the Technical University of Munich. His books include (with John Palfrey) Born Digital: How Children Grow Up in a Digital Age. Viktor Mayer-Schönberger is professor of internet governance and regulation at the University of Oxford. His books include Delete: The Virtue of Forgetting in the Digital Age (Princeton).
by Gary Mintchell | Apr 24, 2024 | Generative AI
I went to the Aras customer conference for the first time this year. Interesting company, good products, innovative customers. But, sorry, I’m hardly shocked that a survey of 835 “executive level experts” say their companies are not prepared to use Artificial Intelligence. We are all still feeling our way along the path toward discovering if there is a use or not. AR and VR are much farther along the hype curve and still haven’t really found a place.
However, you can check out all the details here.
Aras, a leader in product lifecycle management (PLM) and digital thread solutions, announced today findings from its report, “Spotlight on the Future 2024,” highlighting that nearly 80% of industrial companies lack the knowledge or capacity to successfully use artificial intelligence (AI).
Oh, PLM users seem to be the best positioned to benefit. You can pick up a few ideas from my interview with CTO Rob McAveney.
Despite this unpreparedness, 84% of companies expect AI to provide new or better services, while 82% expect an increase in quality. These findings come from Aras’ recent global industry study in which 835 executive-level experts across the United States, Europe, and Japan were surveyed.
“Adapting and modernizing the existing IT landscape can remove barriers and enable companies to reap the benefits of AI,” said Roque Martin, CEO of Aras. Current gaps in the industry according to Aras’ global study, include capacity bottlenecks 79%, lack of knowledge 77%, reliance on isolated IT applications 75%, and existing data quality concerns 70%.
The findings from the report suggest that augmenting product lifecycle management (PLM) with AI leads to improved effectiveness. Some 75% of respondents noted AI’s influences on their PLM strategy, while 2/3 of respondents said that their current PLM platform and data infrastructure is well-prepared for AI technologies.
Martin added, “Companies that are already using a flexible and modern PLM are much better prepared for the challenges of new, data-intensive technologies, leveraging AI to their benefit.”
Study participants rely primarily on datasets such as product data, quality control data, production data, or customer data. Many survey respondents acknowledge their data quality is not enough to achieve their company’s goals. As a result, 51% of respondents are intensifying their efforts to improve production, while 46% are looking at services data, and 45 percent are paying special attention to research and development datasets. These findings show a growing recognition of the important role that high-quality data plays in driving successful AI use within enterprises.
by Gary Mintchell | Apr 23, 2024 | Automation, Sensors, Wireless
Digi International PR has been working overtime sending news to me. This news regards remote monitoring and control.
Digi International introduced Digi Connect Sensor XRT-M designed to deliver reliable and seamless remote monitoring and management. It offers customers both edge and cloud-based control options, with a management platform equipped with an intuitive interface for streamlining data management and enhancing operational efficiency.
Infrastructure monitoring applications include:
- Critical infrastructure and civil engineering projects
- Water/wastewater monitoring and remote management
- Environmental monitoring
- Industrial automation and control
Key Features of Digi Connect Sensor XRT-M include:
- Extended battery life with lightweight cloud connection and dual 14.5 Ah batteries
- Rated for industrial environments with wide temperature range, C1D2 certification with NEMA enclosure
- Device health and sensor data stored in Digi Axess
- Configurable read and report intervals
- Connect to and power multiple sensors
- Compatible with a wide range of industry-standard sensors
- Modbus protocol support
- Global cellular coverage
by Gary Mintchell | Apr 22, 2024 | Operations Management, Workforce
I recently connected with Adrian Heaton, Global Service Manager for ABB Measurement, to discuss the global skills shortage for technical talent in our industry.
A 2021 study conducted by Deloitte and the Manufacturing Institute (MI) predicts that 2.1 million manufacturing positions will go unfulfilled by 2030, partly due to an increasing skills gap.
Following are several points Heaton made during our conversation.
Heaton works in the Measurement and Instrumentation group of ABB focusing on services. He told me that from a service resources point-of-view, finding technicians who want to work in the area and have relevant skills is difficult. In Europe it can take 6-12 months to fill a spot.
He told me, “This is reality—not fake news.” (OK, the US meme has reached Germany, from which we connected.)
It appears to be a pretty global situation that tech colleges are an important source for people with technical skills. I’ve witnessed that in the US for several years. Looks like the same in Europe. He also mentioned that that can be a problem as students graduate in one country where opportunities are not abundant, so they move to other countries to find appropriate jobs.
I asked about something I had heard years ago where the large automation suppliers may have an advantage hiring talented people over smaller companies. He said that sometimes customers are having trouble finding people while the large automation suppliers have brand awareness that helps attract applicants. While marketing is key to attracting talent, training and employment contracts that offer more potential for development are also important.
Training, has always been a strong suit at ABB, but they are standardizing more, making it consistent across lines. E-learning of instrumentation is an adjunct to the classroom. Training is prepared as part of product launches so that all concerned can get thorough updates. “It’s the DNA of the product program.”
Safety is an important part of training at ABB. They are taught to evaluate what is risk. Perhaps instrumentation is low voltage but there are enclosed spaces, and underground places to consider. The global services team puts on as many miles driving to sites as miles around the Earth. So driving safety is another part of the service training DNA. “It’s extensive so that the team gets home at night.”
Heaton confirms my observation that Virtual Reality technology seems best suited for training. But it is very expensive to develop the visuals for VR in order to achieve good industrial training.
I asked about Augmented Reality (AR). ABB has a program Visual Remote Support that allows remote experts to “look over the shoulder” of the on site technician. This program was a success during COVID. People couldn’t travel and customers found the experience quite good. But after COVID, people wanted to meet people face-to-face again, so the program has dropped off some. It remains useful for checking things out prior to a visit such that the traveling technician can bring the right tools and parts.
ABB put together training using AR/VR technologies during COVID. They found customer experience was quite good. There were a number of programs where customers could “send” multiple people to training without travel cost. Students could actually manipulate instruments and valves from their remote locations.
AR remote support faces one major challenge for in oil and gas plants and other plants with hazardous areas—you can’t take an iPad in.
I haven’t had a talk with ABB for quite some time. It’s good to see how a major automation company responds to the skills crisis.
by Gary Mintchell | Apr 19, 2024 | Automation, Sensors, Technology
Computer vision technology and application seemed to have been in a long lull until semiconductor advances of the past few years. Now companies are unleashing new products at an astonishing pace. This is an updated product from Digi International—a company I covered a long time ago who became less interesting to me and now has returned to my view.
Digi International introduced the wireless and highly power-efficient Digi ConnectCore MP25 system-on-module (SOM) at Embedded World 2024. Featuring AI and machine learning capabilities, with an integrated neural processing unit and image signal processor, Digi ConnectCore MP25 is designed for next-generation computer vision applications in critical sectors such as industrial, medical, energy and transportation. The module provides highly reliable wireless connectivity and time-sensitive networking (TSN) making it perfect for smart portable devices and Industry 4.0.
Based on STMicroelectronics’ STM32MP25 MPU processor, Digi ConnectCore MP25 is engineered to streamline application development under demanding requirements to improve efficiency, reduce costs, boost innovation and improve end-customer satisfaction.
With an innovative architecture, Digi ConnectCore MP25 features two Cortex-A35 cores operating at 1.5GHz supplemented by a Cortex-M33 core and a Cortex-M0+. Enhanced with an AI/ML neural processing unit (NPU) delivering 1.35 TOPS and an image signal processor (ISP), the SOM provides accelerated machine learning capabilities for advanced applications. Its comprehensive connectivity options include 802.11ac Wi-Fi 5 and Bluetooth 5.2 wireless technologies, along with seamless cellular integration for expanded possibilities. The ultra-compact Digi SMTplus form-factor (30 x 30 mm) and industrial temperature ranges (-40 to +85 °C), ensure reliability in the harshest environments, making it an excellent choice for a wide variety of IoT applications.
