Data Remains Key Challenge for Artificial Intelligence Projects

70 Percent of Companies Report C-Suite Involvement into AI Projects, with COVID-19 Driving Acceleration of AI Strategies – But Businesses Still Name Data as Key Challenge.

“Artificial Intelligence is neither,” according to my favorite quip. This field of computer science is perhaps the most misunderstood thanks to possibility thinkers like Ray Kurzweil and the many dystopian movies. Nevertheless, AI in its realistic form powers much modern technology. And not only for home artificial listening devices. Business and industry use it often, as well.

Appen Limited, the leading provider of training data for organizations that build effective AI systems at scale, has announced its annual State of AI Report for 2020. The report highlights increasing C-suite involvement and investment in enterprise AI projects as well as data being a key challenge as AI models get more frequent updates in production. The report also reveals the recent acceleration of AI strategies in the wake of the COVID-19 pandemic.

According to the report, nearly 75 percent of businesses now consider AI critical to their success, and AI continues to grow in importance across companies of various sizes and industries. Yet, almost 50 percent of those who responded to the 2020 State of AI survey feel their company is behind on their AI journey, suggesting a critical gap exists between the strategic need and the ability to execute.

“Many organizations have adopted the use of the internet at the core of their processes, and AI is on a similar journey from fringes to core value offering. Increasing investment in AI projects and greater involvement by the C-suite, along with accelerating enterprise adoption in the wake of COVID-19, are clear indicators that AI is core to business success,” said Appen CEO Mark Brayan. “However, most companies are still in the early stages and facing challenges, especially around training data.”

Key Takeaways from the 2020 State of AI Report

The C-Suite is now far more heavily invested and involved in the development of AI projects

Executive visibility and involvement in AI have increased over 30 percent year-over-year, with 71 percent of organizations reporting C-suite involvement in AI projects. What’s more, the percentage of companies investing over $5 million has effectively doubled compared to last year. With this level of executive involvement and increased budgets, ethics, governance, and risk management initiatives have become important topics for technologists building AI.

COVID-19 is not slowing AI Investment

Continuing investment in AI shows that businesses are choosing to spend in times of turbulence. Two-thirds of companies do not expect any negative impact on their AI strategies. Nearly 50 percent of companies have accelerated their AI strategies, 20 percent doing so “significantly,” betting their AI projects will have a positive impact on their organization’s resiliency, efficiency, and innovation.

“COVID-19 has changed everything about the way companies are operating today, but not everyone has adapted in the same way,” added Wilson Pang, CTO at Appen. “The State of AI report shows despite turbulent times, more than two-thirds of respondents do not expect any negative impact from COVID-19 on their AI strategies. Those that are prioritizing AI see the power of digital transformations as a way to improve their resiliency and long-term performance.”

Data remains the key AI challenge

Training data is the key to successful AI, with 3 out of 4 companies updating their models at least quarterly. However, 40 percent of those updating quarterly feel that a lack of data or data management is a challenge.

“Many businesses are still early on their AI journey and they are finding that their data needs span beyond in-house resources when looking for high-quality, annotated training data that drives AI success,” added Pang. “Industry leaders are turning more and more to third-party providers like Appen to help them deploy their AI projects.”

Coding Guidelines For Qualitative and Standardized PLC Programming

Coding Guidelines For Qualitative and Standardized PLC Programming

Can we bring more discipline to PLC programming in industrial control?

Discussion swirls at every gathering of automation professionals about the new generation of engineers entering (we hope) the industry. One thing for sure, the new generation begins with a much deeper computer science background than any before. Will they want to continue to code programmable logic controllers (PLCs) in the same relay graphical representation as their predecessors? Even the Structured Text language popular in Europe and other places can string together like an old BASIC program.

I am guessing they will bring more discipline to the craft of coding industrial control than ever before. Evidently PLCopen, the organization devoted to developing standards for PLC programming, does also. It has just announced a new set of coding guidelines.

The organization notes in its press release, “Software is becoming increasingly responsible, complex, and demanding. This does not come without its challenges. Due to the greater complexity, programs are more difficult to maintain, more time consuming, and potentially therefore more expensive. This is why quality is taking such an important role these days.”

Continuing, they note, “Unlike in other industries, such as that of embedded software and computer science, there has not previously been a dedicated standard for Programmable Logic Controller (PLC) programs. This has meant that programs were not measured against anything and were often of a poor quality. But that’s where the independent association PLCopen has come in and set the standard with the release of their coding guidelines. These guidelines are a set of good practice programming rules for PLCs, which will help to control and enhance programming methods within industrial automation.”

PLCopen, whose mission is to provide industrial control programming solutions, collaborated with members from a number of companies in different industries to create these coding guidelines. These companies include PLC vendors such as Phoenix Systems, Siemens, and Omron, to software vendors such as Itris Automation and CoDeSys, and educational institutions such as RWTH Aachen. These guidelines were inspired by some pre-existing standards from other domains such as IEC 61131-3, JSF++ coding standard and MISRA-C, and they are the product of three years of work by the working group. PLCopen’s reference standard can be used for testing the quality of all PLC codes, independent of brand and industry.

PLCopen’s coding guidelines are made up of 64 rules, which cover the naming, comments and structure of the code. By following these guidelines, the quality of the code will be improved and there will be greater consistency amongst developers. This will result in greater efficiency, as better readability means a faster debug time, and a program that is easier to maintain. This then results in lower costs as less time is required in order to maintain the program, and the maintenance should be easy enough for both an internal or external programmer as the code will be more straightforward. If the original developer fails to follow certain guidelines when creating a program, this could obstruct other developers and maintenance teams when working with the code during the product lifecycle, thus creating delays and additional costs.

In safety-critical industries, there is the standard IEC 61508 which in 2011 was also extended to PLCS. However, as quality is becoming an ever more important factor across the board, as programs become bigger and more complex, it is generally good practice to follow a set of rules or a standard in all industries. PLCopen’s coding guidelines suggest a standard that can be used across all industries to greatly improve the quality of the code and, as a result, to help companies save time and money. The introduction of such a standard will allow PLC programs to be verified not only from the simple functionality perspective but also from a coding perspective by confirming that good practice programming rules have been followed in their creation. Consistency across PLC programs can only be achieved through the respect of a global corporate or industrial standard, with PLCopen now being the de facto standard in the automation industry.

With quality playing a greater role in industry and with companies always looking for cost saving methods, the answer is to use some sort of standard or set of rules in order to meet these goals. PLCopen have created this standard to improve quality and consistency across PLC programs and so that individual industries and companies don’t have to go to the effort of creating a set of rules themselves. In addition to the internal benefits, this standard will also allow companies to enforce their quality requirements on suppliers, software contractors and system integrators. The only issue for now is that the process for verifying these rules is done manually by most users as they are unaware that some tools are available to do this automatically. But overall, following a standard such as the one proposed by PLCopen, will greatly improve the quality of the program and will save time and money throughout the whole duration of the product lifecycle.

The PLCopen coding guidelines v1.0 are available to download for free from the PLCopen website.

Coding Guidelines For Qualitative and Standardized PLC Programming

ABB Searches Silicon Valley for Digitalization Leader

ABB chief digital officerOnce again, an industrial corporation striving for digitalization solutions looks to Silicon Valley. Industrial and automation companies continue to search outside the industry for talent. Often, they just don’t work out. The pace of change in industrial is so much slower and bureaucracies are so entrenched that they get frustrated and leave. This one from ABB is interesting. Will he be able to build a software division to rival GE? We’ll keep watching.

ABB has appointed Guido Jouret as Chief Digital Officer, reporting to ABB’s CEO Ulrich Spiesshofer, effective October 1, 2016. Labeled a “pioneer of the Internet of Things” by ABB, Jouret will lead the next level of development and deployment of ABB’s digital solutions for customers globally and across all businesses.

Jouret is a citizen of both the United States and Belgium, with long experience in Silicon Valley. He served for 20 years at Cisco after obtaining a PhD in Computer Science. His most recent role at Cisco was as General Manager of the Internet of Things division, which aims to connect billions of devices to the internet across a wide range of industries.

Prior to that, he was Chief Technology Officer and General Manager of Cisco’s Emerging Technologies Group, a unit responsible for incubating new businesses. Under his leadership, the team created nine new start-ups, including those, which became Cisco’s TelePresence and Internet of Things groups. He has lived in 12 countries including France, Singapore and the United States. Additionally, he worked for Cisco as IT Director for the Europe, Middle East and Africa region and in the Internet Business Solutions Group.

Jouret left Cisco in 2014 to join Envision Energy, where he led the software products business, including a platform for the emerging energy internet. Jouret has been Chief Technical Officer at Nokia Technologies since April 2015, and recently drove a major acquisition in the digital health sector.

“Guido Jouret is a proven leader in the digital revolution, with a solid track record in creating and growing new businesses as well as in the digital transformation of mature businesses. His global digitalization experience spans across companies in utilities, industry and transport & infrastructure,” said CEO Ulrich Spiesshofer. “With his customer and innovation focus, and his extensive international technology business experience, he is the ideal person to take the software-led differentiation of our offerings to the next level and to cement ABB’s pioneering technology leadership, building on our more than 40 years’ history in embedded, control and applications software and our strong position in the Internet of Things, Services and People.”

Jouret will work from ABB’s Silicon Valley and Zurich offices.

He obtained his PhD in Computer Science from the Imperial College of Science, Technology and Medicine in the UK in 1991, and has a bachelor’s degree in Electrical Engineering from Worcester Polytechnic Institute in the US. He is married and has two children.

Coding Guidelines For Qualitative and Standardized PLC Programming

Never Stop Learning

How many careers have you had?

Some may have chemical engineering or computer science degrees and are still doing process control or computer networking. But many of us have wound up with a circuitous path to where we are now. In a different industry. Or doing sales or management rather than pure engineering.

The key for career flexibility and advancement is continuous learning.

For a very long time, I’ve been concerned with the prevailing “wisdom” that education exists solely for vocational enhancement.

Here is a voice from the Silicon Valley venture capital community issuing a warning much as I would. In Hard-Core Career Advice for a 13-year-old, James Altucher notes, “[My experience] shows that school is too focused on ‘education leads to a job.’ This is not true anymore. “

He continues, “The reality is the average person has 14 different careers in their lives and the average multi-millionaire has seven different sources of income. So anything that is ‘one-job focused’ will create a generation of kids that will learn the hard way that life doesn’t work like that.”

I have always believed that education is necessary for personal growth.

There are more of my thoughts on this topic in my (mostly) weekly newsletter that went out today.

I am typing this article outside the press room on the pool deck of the Grand Hyatt in San Diego. Press room? Yes, I’m at Rockwell Automation TechED. I find myself needing to cut back on the amount of travel. In a couple of weeks, I’ll have to report on both Honeywell and Siemens from the reports of my friends. Yes, they are the same week in different parts of the country. I just couldn’t make them. But this week in San Diego worked.

Plus, this is my 9th one of these, I think. Formerly RSTechED (they like the capital d–that makes it a logo rather than text), it is now dubbed Rockwell Automation TechED. The reason is there is an expansion of training opportunities beyond the initial HMI/SCADA and programming software. It now includes information systems, new commercial technologies, and discussion of new hardware products.

Attending these events is one way professionals participate in continuous learning.

Connected Enterprise

Rockwell has maintained a consistent theme for many years–the Connected Enterprise. It is still building upon that vision. Cisco, Panduit and Microsoft remain as top-tier partners. EtherNet/IP, the Industrial IP Consortium, and mobility remain as foundations.

There will be more to contemplate this week as I have four defined interviews and many other opportunities. The connected enterprise really is a vision beyond just the Industrial Internet of Things. And Manufacturing 4.0 remains a German initiative mostly targeting Germany’s strength in machine building. I’ve been removed from the US “Smart Manufacturing” circle, but I don’t see it really have a huge impact in the market.

But smart, connected devices, machines, lines, plants, and enterprises still point to the future of manufacturing.

Follow this blog

Get a weekly email of all new posts.