SmartIndustry – What would your plant look like after a 20-year nap

If you were a production engineer who fell asleep at your laptop and woke up 20 years later, what would you behold?

That classic time-traveler, Rip Van Winkle, slept through the Revolutionary War and awakened to an unfamiliar country after his 20-year slumber. Would your operations, circa 2038, be just as foreign?

It’s not such an outlandish scenario, given the revolutionary change that is afoot in today’s industrial world. Digital technologies are redefining how some companies understand, manage and staff their industrial operations. Yet, for so many others, terms like Industry 4.0 and digital transformation are still abstract concepts.That classic time-traveler, Rip Van Winkle, slept through the Revolutionary War and awakened to an unfamiliar country after his 20-year slumber. Would your operations, circa 2038, be just as foreign?

So, let’s look 20 years into the future to see how today’s smarter controllers and other digital technologies could reshape operations in different industries.

Food & beverage: Predicting an end to unplanned downtime?

Your packaged-foods plant has held up surprisingly well after 20 years. But there are noticeable changes.

For instance, you see fewer operator workstations on the production floor. Instead, you see the plant manager and an engineer looking at a tablet, while a robot replaces a device on the production line.

“What broke?” you ask them.

“Nothing,” the plant manager responds. “We have another week and two days before the pump goes down on the mixing tank. Our system predicted a failure last month and put an order in for the replacement part. We’re putting the new pump in today during our scheduled maintenance.”

You watch as the plant manager pulls up the plant analytics on his tablet and quickly drills down into the failing pump, showing its expected failure date. He then pulls up a display of all plant devices that need to be serviced within the next month.

Having just arrived in the future, knowing nothing about it, you’re amazed by how analytics software can now see into the future with enough clarity to predict downtime issues. Twenty years ago, you think, the plant was still doing calendar-based maintenance. Workers couldn’t predict anything, and they often found themselves fixing the same re-occurring problems.

You’re also surprised by just how easily information is now accessed. For example, the plant manager can now search his operations for answers—such as machine health or performance KPIs—as if he were searching the internet. He can even verbally ask his tablet questions, like how long the next batch will take to produce, and the software speaks back with answers.

Automotive: Robots save your back (and more)

Who knew robots could be so intuitive—and so helpful?

As you stumble out of your slumber and onto the plant floor, you are amazed by how seamlessly and safely robots are now woven into production processes.

Approaching the assembly area, you don’t see a gate around the robots to keep people at a safe distance. Instead, as you walk closer, the robot slows down and then comes to a stop as you stand next to it.

But it’s not just the proximity of the robots—it’s what they’re doing. Before drifting into your decades-long dream, you were worried about the safety implications of the heavy lifting being done by the plant’s older workers. And you worried that the plant wasn’t going to find skilled talent to replace those workers when they retired.

Now, robots do all the heavy lifting. This helps keep production running—even through labor or skills shortages. It also improves ergonomics for workers, reducing their exposure to repetitive strains.

All this is made possible by a combination of safety technologies, including robots, sensors and controllers.

Life sciences: A multi-modal mobile makeover

Is this even the same production facility? You look around to see that the sea of fixed, stainless-steel

biopharma machinery has been replaced by modular, mobile equipment.

It looks like a coordinated dance: workers move equipment throughout the facility and reposition it with simple, plug-and-play connectivity. Instead of large batch runs, you see workers making smaller, more targeted batches or even single drugs that are personalized to one person’s physiology.

You intently follow one operator as she pushes a piece of mobile equipment from one production area to another and then connects it to a docking station. The control system automatically detects the equipment connection, identifies its IP address and confirms that it’s in the right location. The operator then uses a tablet to guide her through connecting the right tubes to the appropriate points on the transfer panel.

Making all this happen is a modern distributed control system (DCS) with an open and unmodified Ethernet backbone. The DCS can confirm that the right mobile equipment is connected—and only allows control when equipment is in the right location. It also allows workers to scan materials to confirm they’re used with the right equipment.

Thin-client technology makes mobile visualization possible by recognizing an operator’s location and allowing access to screens and applications that are relevant to that location. And asset-management software securely and centrally manages production.

Make this future a reality

These future factory scenarios aren’t simply theory or fantasy. They’re the inevitable outcomes of adopting the control and information technologies that are available today as part of an Industry 4.0 strategy. And they’re closer than you might think.

For example, new controllers with onboard computing can help you tie production more closely to customer orders. And new analytics software enables facility workers to quickly and easily access analytics that aren’t displayed on their dashboard. They can even use natural-language search to quickly and easily find information in their plant.

Forward-thinking companies are already beginning to make these future factories a reality, gaining all the competitive advantages that comes with it. This is why every organization should be thinking about how their Industry 4.0 strategy will help them compete, not just today, but in the decades to come.

https://www.smartindustry.com/blog/smart-industry-connect/rip-van-winkle-and-digital-transformation/

TechTarget – Deep learning use cases aren’t limited to big tech companies

Industries that are not traditionally technology-driven are starting to find ways to use deep learning, proving the tools aren’t just for large tech companies.

First of all, it’s not the kind of project an advertising firm typically takes on. And it was also unclear if the growers had the technical foundation on which to build this kind of deep learningcapability. But speaking at the Spark + AI Summit in San Francisco, the company’s vice president of engineering, Sam Saha, said he felt it was worth trying as a deep learning use case.

“There are lots of businesses out there and they don’t have the talent pool [to use AI],” he said. “It becomes a vicious cycle where the innovation isn’t flowing downstream. At some point, you have to just get up and try.”

Deep learning not just for big tech companies

Most of the deep learning use cases we’ve seen in recent years have come from large tech players like Amazon, Facebook and Google. This has created a perception that, while a powerful tool, deep learning is just for the leading companies, with limited applicability in more general enterprise settings. But some are out to prove this perception wrong.

Saha said his company approached the project as a proof of concept that would demonstrate how less tech-savvy companies can implement AI tools.

His team built an iPhone app that enables cranberry growers to take pictures of suspected pests. Initial images were labeled by employees at Ocean Spray, a major buyer of cranberries that has an interest in the health of bogs. Then, Laughlin Constable data engineers built a deep learning model using TensorFlow to identify specific varieties of pests in the farmers’ pictures.

Saha said it’s a good example of how AI tools like deep learning can be used by organizations outside the technological hubs of Silicon Valley and the I-95 corridor. It also shows the power of open source tools.

The deep learning model used in the project was already built and available in TensorFlow; it just needed to be trained. But Saha said he doesn’t see a lot of people in enterprises today taking this deep learning use case approach. Instead, he sees a belief that if you aren’t writing models from scratch, you aren’t doing real AI, which is self-defeating and impedes AI adoption at businesses that might benefit from it.

“On the business side, people are always worried about whether we actually need AI,” Saha said. “On the engineering side, they say we’re not Google. We’re not Amazon. Where do we start? Before they even attempt to solve a problem, they give up.”

Building AI in the construction industry

It’s hard to get much farther from AI than pouring concrete and laying pipes, but at the summit, a senior data scientist at construction company Bechtel Corp., Evann Smith, described the company’s deep learning use case, which is aimed at optimizing construction planning.

Smith is currently leading a project to use reinforcement learning models similar to those used by AlphaGo, the AI model that beat human champions of the game Go in 2016, to find the fastest route to build projects. The model runs step-by-step simulations of projects, testing out sequences of laying concrete and installing pipe to find the optimal sequence.

There are a number of characteristics unique to construction that have historically left the industry less reliant on technology than others. For one, Smith said, each project is unique, which means there’s essentially no set of training data from past projects that can be used for machine learning jobs like the one she is running. That’s one reason why she’s using reinforcement learning, in which the simulations essentially become the training data set.

Also, the Agile development method, which has become the leading paradigm for building AI applications at most companies, doesn’t really apply to construction, in which projects are completed sequentially rather than iteratively.

Still, Smith said she thinks the potential for bringing deep learning use cases to the construction industry is huge, and that Bechtel is just starting to explore that promise.

“The goal is to take our industry knowledge and pair it with this deep learning to move the industry forward,” she said. “The potential to optimize is really major.”

https://searchenterpriseai.techtarget.com/feature/Deep-learning-use-cases-arent-limited-to-big-tech-companies

HBR – Why We Need to Update Financial Reporting for the Digital Era

The market caps of just four companies, Apple, Alphabet, Amazon, and Microsoft, now exceed $3 trillion. Their combined assets of $944 billion are an order of magnitude lower than the combined assets of $7,700 billion of the largest 3,177 companies in 1986, when the aggregate market capitalization reached $3 trillion for the first time. In our recent HBR article, we argued that financial statements fail to capture the value created by modern digital companies. Since then, we interviewed several chief financial officers (CFOs) of leading technology companies and senior analysts of investment banks who follow technology companies. We asked: (i) what makes the valuation of digital companies more challenging?; and (ii)  how can digital firms improve their financial reports to communicate sources of value creation in their businesses? We distilled seven key insights from those discussions.  Some of these ideas contradict traditional financial thinking whereas others seem highly controversial or pessimistic.

Financial capital is assumed to be virtually unlimited, while certain types of human capital are in short supply.

Business students have traditionally considered net present value, payback period, and hurdle rates as necessary tools to determine which project to select. These criteria assume a limited supply of financial capital and that prudent allocation of financial capital determines a firm’s success. Digital companies, however, consider scientists’ and software workers’ and product development teams’ time to be the company’s most valuable resource. They believe that they can always raise financial capital to meet their funding shortfall or use company stock or options to pay for acquisitions and employee wages. The CEO’s principal aim therefore is not necessarily to judiciously allocate financial capital but to allocate precious scientific and human resources to the most promising projects and to pull back and redeploy those resources in a timely manner when the prospects of specific projects dim.

Risk is now considered a feature, not a bug.

Traditional valuation models consider risk to be an undesirable feature. Digital companies, in contrast, chase risky projects that have lottery-like payoffs. An idea with uncertain prospects but with at least some conceivable chance of reaching a billion dollars in revenue is considered far more valuable than a project with net present value of few hundred million dollars but no chance of massive upside. In light of this, an employee is evaluated not based on what she contributed to the company’s bottom line, but whether she identified a new, breakthrough idea.

This notion, that risk is a desirable feature, can seem like sacrilege to anyone who’s taken an introductory finance course. It’s unlikely that investors’ risk aversion has fundamentally changed. However, many investors seem to have concluded that the most successful companies with tens of billions of dollars of valuation today could never have justified their valuation at the start of their operation based on discounted cash flow. So, investors, and therefore managers, might be adjusting their approach to risk accordingly.

Investors are paying more attention to ideas and options than to earnings.

Business students are taught to value a company based on the discounted amounts of future cash flows or earnings. That concept is becoming almost impossible to apply to emerging companies that are run as a portfolio of ideas and projects, each with uncertain lottery-like payoffs. CFOs of these companies themselves admit that they cannot justify their market capitalizations based on traditional metrics. They conjecture that their market values might be the sum of the option values of the projects undertaken, a sum of best-case scenario payoffs. One CFO said that her valuation should be considered on a per idea basis instead of a per earnings multiple.

In theory, options valuation should be able to handle this problem of valuing firms with lottery-like payoffs. In practice, we have yet to see a model that can justify, for instance, Amazon’s market capitalization. It’s possible that companies like those are overvalued. It’s also possible that we simply don’t know how to estimate the right parameters to make an options-based valuation work.

As digital technology becomes more pervasive, more and more companies will present this sort of valuation challenge. Given that even sophisticated investors cannot estimate the value of these companies, CFOs question the ability of a day trader to value a digital company. Therefore, companies see little value in disclosing the details of their current and planned projects in their financial disclosures, even if those disclosures can reduce the information asymmetry between investors and managers. Given the bull market in digital stocks, CFOs believe that they have no incentives to provide any additional information beyond mandatory SEC disclosures, which they consider excessive, tedious, and uninformative and might invite unnecessary scrutiny and litigation.

Corporate venturing is becoming more important.

Many traditional companies realize the potential for disruptions in their business models from the likes of Alphabet and Amazon. However, they do not possess the infrastructure or talent pool to ward off potential competition. Furthermore, the operating managers cannot take their eyes off day-to-day operations to focus on innovation. Traditional companies therefore rely on two strategies. First, they create a new venture capital (VC) arm within the existing organizational set up. That VC arm is given relatively unconstrained financial capital to invest in innovation and disruptive ideas. Second, companies perform “acquihires,” — that is, the buying of a company primarily for its engineering and product design talent, instead of for its revenues or profits. Both strategies, however, create cultural incompatibility within the organization. For example, the legacy part of the organization is subject to tight fiscal discipline while the VC arm is given pots of money to bet on new ideas.

CFOs realize the growing limitations of the current financial reporting model. They are, however, extremely pessimistic about whether the model can be fixed within the current regulatory regime. One CFO commented that standard setters enjoy monopoly power and have no incentives to change their methods to be more responsive to investors. Another CFO mentioned that it will take a full-blown crisis, such as the 2000 dot-com meltdown, to force substantive changes in the standard setting process. In the meanwhile, companies increasingly resort to provision of proforma and non-GAAP reports, even though this practice is looked down upon by the SEC and is opportunistically misused by a few companies.

Analysts increasingly rely on non-GAAP metrics.

As firms become increasingly difficult to value and more and more companies report negative earnings, analysts perform multiple adjustments to recreate companies’ financials in their internal assessments. For example, they capitalize a part of R&D expenditures that can enhance firm’s future competitive ability and deduct a part of capital investments that merely maintain firms’ competitive ability.

Sadly, accounting is no longer considered a value-added function.

This is an outcome of the growing divergence between what companies consider as value-creating metrics and those reported as profits in the GAAP. Many CFOs consider financial reporting to be an exercise in mere regulatory compliance and find the resources spent on audits and financial reporting to be a waste of shareholder money. They consider the calculation of GAAP-based profitability to be more of a hinderance and distraction to their internal resource allocation decisions. One CFO commented that they now avoid inviting company accountants to their strategy meetings, while another said that CPA certification is considered a disqualification for a top finance position.

It’s clear to us from our research and from these interviews that the time has come for investor bodies and companies to rethink the financial reporting model from scratch. For instance, standard-setters might want to encourage disclosures related to (i) value per customer; (ii) earnings or revenue outcomes or other specific metrics related to specific projects in progress; and (iii) data on how the R&D and software talent of digital firms is being deployed.  Relying on firms’ voluntary initiatives is unlikely to work because executives told us time and again that they will not disclose sensitive information, unless their competition is forced to do the same.

https://hbr.org/2018/06/why-we-need-to-update-financial-reporting-for-the-digital-era

DesignNews : Small Companies Make the Leap to Advanced Manufacturing

As smart manufacturing tools become more intelligent and easier to use, they’re becoming affordable for small manufacturers.

As manufacturers adopt more technology and move to highly sophisticated automation systems, much of the change does not involve edgy tech breakthroughs. Rather, the shift is a matter of existing technology moving into the hands of smaller manufacturers. Predictive maintenance, IoT, connected systems, moving data from OT to IT and back—all of this has been in the hands of major manufacturers, such as Boeing, GE, and Ford, for years. Now, it’s moving into the hands of the shop down the street.

Appliance for connecting OT and IT systems
The Data Commander gateway appliance moves data from OT to IT without the need for original programming. (Image source: Ellitek)

“Moving to the factory of the future is an evolution rather than a revolution. Manufacturing has been able to collect a lot of data for years, but now we’re seeing that it’s becoming available to small- and mid-size manufacturers,” Keary Donovan, market development manager at elliTek, told Design News. “Whether it’s zero-defect initiatives or ERP systems, smaller companies are beginning to get the benefits.”

Atlantic Design & Manufacturing is the annual must-attend trade show for discovering the latest in design engineering. From prototyping to full-scale production, the show floor will keep you up to speed on the innovations transforming the industry.

While the devices and sensors associated with smart manufacturing technology are becoming less expensive, the greater impact may be that this equipment has become easier to deploy. “The products are getting simpler to use and they’re going to get more dispersed,” said Donovan. “That can include analytics, lot size of one, or mass customization. These are all becoming available to smaller manufacturers.”

The Intelligence Is Built into the Device

Part of the affordability of new manufacturing technology comes from the ability to deploy it using configuration instead of original programming. “We saw a bunch of controls engineers witness a collaborative robot presentation. The sales person was proud of the robot’s ability to open and close a door,” said Donovan. “All of the controls engineers flipped out negatively. ‘We can program that,’ they said. Yet they couldn’t understand the fact that nobody wants to be dependent on them for all that.”

Donovan points to the appliance elliTek makes as an example of intelligence that is built into the device. The Data Commander connects factory automation data to database servers. It’s a high-speed, two-way gateway connection that protects data integrity and isolates OT from IT networks. “Our product doesn’t require programming. All of the drivers are included in the product,” said Donovan. “Instead of needing a PLC or Java engineer, you can point and click through a free interface to map the fields of the machine controls like PLCs to the database.”

While smart manufacturing technology is not necessarily cheap, the overall cost of ownership is significantly lower when it doesn’t require expensive outside experts to set it up and run it. “Getting end-to-end (production connectivity) still costs up to $12,000. But you own it when you’re done. You don’t need an integrator,” said Donovan. “If you have a collaborative robot, you’re not calling a controls engineer to teach it a new function. You’re not calling a system integrator every time the production run changes. You just need a person who understands manufacturing processes.”

Using Tech to Empower Users

Smart manufacturing mostly involves obtaining and using data to improve production efficiency. The ability to get the data is the key to that efficiency. “Everyone wants the data, but it’s shocking how few people have access to it,” said Donovan. “Beyond ease of use for efficiency efforts, the dedicated firmware on our appliance is a ‘black box’ solution that eliminates FDA software validation and its engineering costs for data exchange.”

Donovan noted that those who are developing smart manufacturing tools need to make ease-of-use a priority. “The goal is to make it easy, to make sure people are not distracted by analytics. We need to make sure the data enhances their job rather than getting in front of their job,” said Donovan. “We challenge ourselves to make that part of our product line. Manufacturing process engineers and production managers don’t want to have to learn programming languages. Modems are an internet gateway appliance. Users are are only expected to be able to plug it in, not how to program it.”

He explained that those producing the new technology need to make sure they’re solving a problem rather than creating a new problem. “During the Mac vs. PC days, Steve Jobs explained that you’re not supposed to just put a tool in front of the problem. You don’t want the tool to become its own problem,” said Donovan. “I don’t want to protect my machine from latency issues caused by ITs security patches or become a C# or Java programmer to solve my production problem

https://www.designnews.com/automation-motion-control/small-companies-make-leap-advanced-manufacturing/73380363458870

Scroll to top