How New Technology Helps in a Crisis
The shock that the coronavirus has delivered to the economy cannot be overstated. During and after the crisis, Minnesota-based businesses will be relying even more heavily on technology.
Leveraging technology is one way that businesses can run their operations more efficiently. Plus, technological solutions often are demanded by customers. Twin Cities Business interviewed business leaders to learn how they are taking advantage of technology as well as what innovations are on the horizon.
Artificial intelligence is one of the most tech-forward ways that businesses are boosting efficiency, though not the only one, by any means. But the goals of incorporating newer, digitally powered tools are similar: greater company efficiency and more satisfied customers.
“Most companies have gone through the phase of investing in technology for efficiency,” says Ravi Bapna, Curtis L. Carlson chair in business analytics and information systems at the University of Minnesota Carlson School of Management. “Now they’re looking at investing in technology for innovating.” ERP (enterprise resource planning), CRM (customer relationship management), and communications platforms from simple email to Slack and teleconferencing are tools in many company workshops. Now, more businesses are looking to newer technologies to reduce operational costs, improve their products, boost innovation, and better understand their customers and their desires.
“The most significant technology that is before us and that in many ways is a game changer for most businesses comes in the form of machine learning, advanced analytics, and AI,” Bapna notes. “Companies have invested so much over the years in putting in all of these IT systems that generate tons of data.” At the same time, the rise of mobile technology means that consumers are using smartphones and other devices that are leaving almost limitless data trails. “We live in a world where businesses find themselves drowning in data,” Bapna notes. But companies can organize that data, analyze it, and gain insights to improve their operations.
AI on the farm
AI is being applied in numerous sectors. Four years ago, Arden Hills-based agribusiness cooperative Land O’Lakes began developing AI-based solutions to help its farmer-members improve corn and soybean yields. Those innovations helped farmers reduce their use of weed control products, fertilizer, and other costly inputs.
They have plenty of variables to consider. “You need to decide what hybrid to use, the practices you need to leverage to make sure that you’re getting more out of less input,” says Chakra Sankaraiah, Land O’Lakes director of digital technology innovation. Farmers also need to understand the quality of the crop during the year, in the area where the farm is located. Land O’Lakes developed AI-generated prediction tools that can forecast yields based on a variety of data. This allows farmers “to apply inputs selectively and efficiently, and not uniformly all across the field,” Sankaraiah says.
The data that Land O’Lakes accesses includes records of how certain seeds performed at different times of the season in multiple locations. Data also is available on how the seed performs in various parts of the field—near edges and roads, for instance. Other information sources for these machine learning tools include data gathered from remote satellite imagery and data recorded about soil and topography conditions. Data is also gleaned from plant samples to determine nitrogen and other nutrition levels. Land O’Lakes leverages its own test plots—controlled environments—to identify factors influencing various hybrids’ performance, Sankaraiah says.
While large companies like Land O’Lakes have plenty of data to fuel their AI, Bapna says that smaller firms also can use machine learning. “A small business can easily use off-the-shelf deep learning software,” he says. Examples include Microsoft Azure cloud-based machine learning tools and Google’s TensorFlow open-source AI platform. “The only constraint is that [smaller companies] may not have the level of historical data that large companies might have,” Bapna says. Ambitious startups and other small companies should make sure that “they’re collecting data, storing it, and processing it in ways that allow it to be used for AI/machine learning applications,” he says.
On the production floor
Land O’Lakes also manufactures dairy products and applies AI technology in its production facilities to detect variations in product quality that require equipment adjustments. AI is being used in similar ways by other types of production facilities.
Manufacturers also are making use of other types of tech. Before the coronavirus upended the economy, many metro-area manufacturers were struggling to find enough skilled and semi-skilled workers. To address the labor shortage, manufacturers have been trying to utilize technology as much as they can, says E. J. Daigle, dean of robotics and manufacturing technology at Dunwoody College of Technology in Minneapolis.
Several companies that rely on machining in their production processes are buying collaborative robots “like you wouldn’t believe,” Daigle says. He describes them as a newer, more flexible iteration of industrial robots, coming from vendors that include Danish firm Universal Robots and Japan-based FANUC. A traditional robot “has to be protected from any human interaction,” Daigle says. That requires production facilities to install expensive protective barriers and systems around the robot.
Land O’Lakes developed AI-generated prediction tools that can forecast yields based on a variety of data. Farmers then apply inputs selectively, which allows them to save money and promote sustainability.
Read more from this issue
In contrast, a collaborative robotic system can work alongside humans. The collaborative robots, frequently called cobots, are typically smaller than traditional industrial robots. While the cobot can serve as an aide to a human worker, it also has detection features that help protect humans. That’s because the robot is smart enough to not move in human presence if the worker gets too close to the robot, Daigle adds. The robotic system will shut down to prevent human injuries. The collaborative robot is less expensive than a traditional industrial robot. As a result, “what was once a $300,000 sell or a $1 million sell might now be a $40,000 robot,” Daigle says.
Traditional robots are stationary and perform specific tasks, while collaborative robots can be moved to different parts of the business and can be readily programmed to take on new assignments.
What makes collaborative robotics feasible is that CNC machining is a fundamental component of much modern manufacturing. “We’re pretty good at writing code to get the machine to run,” Daigle says. A collaborative robot can be programmed to load raw metal or plastic material into a CNC machine, run the machine, remove the finished product, inspect it under a machine vision system, and then package it.
CNC machining long has been one of the production processes at Maple Plain-based Protolabs, which specializes in rapid prototyping and short-run production of custom parts for a variety of sectors, including medical technology, electronics, automotive, and consumer products. Another is 3D printing, which currently represents about 13 percent of Protolabs’ sales revenue. “It is growing faster than our other processes,” Protolabs president and CEO Vicki Holt says.
It’s a process that continues to generate new innovations. One such technology, which Protolabs has begun evaluating, is digital light synthesis (DLS), a new plastic 3D-printing process from California-based Carbon3D Inc. One of Carbon DLS’ advantages, Holt says, is that it can be used with a wide variety of thermoplastic materials. “It tends to fit where you have a larger number of parts that you’re trying to make of a single geometry, and when you can optimize how those parts are built,” she says.
Each 3D-printing process has its own distinctive characteristics and cost profiles, “so you always want to put your project into the type of technology that is best for the customer’s particular need,” Holt says. Three years ago, Protolabs added HP’s Multi Jet Fusion plastic printing technology to its toolkit. According to Holt, Multi Jet Fusion offers “greater speed, better surface characteristics, and better material properties” than some of the other 3D processes. Her company is exploring the addition of new printable materials, including polypropylene.
Protolabs also has begun to “upgrade and enhance the technology stack in our architecture,” Holt says. That means she envisions “bringing new information and tools to our customers very rapidly through an integrated, end-to-end digital thread model, from e-commerce all the way to fulfillment.” This technology platform, Holt adds, “will offer more ways that customers can manage their projects and their information. And as the digital thread flows through our process, it will provide opportunities for us to drive the productivity” of Protolabs employees, whether in customer-facing or production roles. A customer’s order information “can flow through our system with less human touch,” she says.
Cutting-edge technology tools have also entered a perhaps surprising realm—law firms.
Digitally driven efficiency is crucial for large firms like Minneapolis-based Dorsey & Whitney, which has 1,190 employees in 19 offices throughout the U.S., Canada, and China.
“Like any organization, we want to operate as efficiently as possible,” says Caroline Boudreau Sweeney, Dorsey’s director of knowledge management and innovation. Law firms are facing cost pressures, particularly in practice areas that have been increasingly “commoditized.” At the same time, more clients are expecting their legal counsel to be technologically up to speed. “Adopting technologies and developing tools are differentiators in the market,” Sweeney says. “The more that firms are trying to differentiate, the more critical it is that you stay on top of the technology in order to benefit your clients.”
One technological solution that Sweeney says Dorsey has been building out is a contract review and management system. This platform allows clients to upload contracts and related information. The system automatically routes it to a review team, which submits the contract for AI analysis and returns the results to the platform. The attorney reviews the AI analysis of the contract. The contract is then automatically routed to a senior attorney, who performs the final review and sends it, via the platform, to the client for a signature.
Dorsey also has been using AI-driven tools to speed up legal research and analysis. These processes include electronic discovery, or e-discovery, which gathers electronically formatted information and evidence to build cases. In the past year, the firm has added AI technology for contract language identification. Traditionally, a firm has attorneys or paralegals go through each contract and identify specific legal language—for instance, wording related to termination clauses or liability.
“Sometimes that language is clearly called out; other times, it’s embedded in another clause and isn’t labeled exactly the way you’d anticipate,” Sweeney notes. Using software developed by Toronto-based Kira Systems, Dorsey attorneys and staff can identify and mark this language “literally in minutes,” she says. “This frees up the associates that were doing that to take on higher-level work on other aspects of a deal or project.”
Dorsey is looking to add other technologies to automate its document production processes. One such tool would help attorneys automate drafting motions, in large part by allowing quicker integration of frequently used phrases and clauses. “This would take some of the repetitive work out of the motion drafting and allow the associates that are doing that work to utilize or develop their skills at a higher level,” Sweeney says.
In health care, technology is used in numerous ways, many of which have become familiar to patients. Other uses are less known and more specialized.
Three years ago, the Chu Vision Institute in Bloomington became the first ophthalmology practice in Minnesota to offer SMILE laser-vision correction surgery. “This is essentially a minimally invasive form of Lasik,” says Y. Ralph Chu, founder and CEO. In other words, SMILE creates a smaller incision in the eye compared to Lasik.
Chu has also incorporated into his practice “better ways of determining whether a patient is a good candidate for eye surgery.” He notes that “one of the biggest advances in our field over the last several years has been in devices that allow us to screen the eye.” These digitally driven tools, called topographers, can deliver a three-dimensional view of the eye to the surgeon.
A related technology is a laser-based diagnostic tool called optical coherent tomography (OCT). “It’s almost like a CT scan of the eye,” Chu says. Using OCT technology, “we’re able with a two- to five-minute scan to see different cell layers of the retina. So we’re able to diagnose conditions like glaucoma, macular degeneration, and other aging conditions of the eye sooner and more quickly than we were able to do before.”
Before it closed to battle the coronavirus, the Nickelodeon Universe amusement park’s staffing needs resembled the huge swings of a roller coaster. Having too many employees on hand can be costly; not enough could result in long lines of fidgety kids and cranky parents.
Not so long ago, Mall of America (MOA) addressed this challenge using historical data to predict ridership and staff needs for the upcoming year. “That was a very lengthy, time-consuming process,” notes Phil McDonald, MOA’s data analytics manager.
In the fall of 2017, McDonald and his IT comrades at MOA began exploring a more efficient approach. The technology-driven techniques they put in place fall under several interrelated categories, including data analytics, artificial intelligence, machine learning, and deep learning. Very simply, this digital tech “family” uses the abundance of data a business can access to quickly derive useful insights about processes and customers. The more data these technologies collect, the better they can identify potential and existing problems, solutions, and opportunities. They do so by “teaching” themselves to improve their algorithms.
Using AI technology, MOA dived deeper into historical ridership data while accounting for other factors, such as weather patterns and the times and days students are likely to be out of school. With AI tools, “it takes a couple of hours at most to put staffing estimates together,” McDonald says. This helps MOA save money on the administrative side. It also can help the mall predict revenue from Nickelodeon Universe operations.
“When you think of AI and machine learning, it’s such a broad field,” McDonald notes. “For us, we focus on: How can we use this to help us improve the guest experience? How can we use this to help us make more informed decisions about how we operate our business?”
In sum, newer technologies allow businesses of all kinds to move faster. At MOA, for instance, the mall is using sensors and digital communications technology to quickly identify when an escalator is out of service—and, just as quickly, direct guests to an alternate route. Mall visitors accessing MOA’s digital directory via kiosks or on their smart devices to get to a certain store, restaurant, or attraction are provided directions that route them away from the out-of-service escalator.
“That results in a better experience for guests, making sure that it’s easy for them to get to the place they’re looking for,” MOA’s McDonald says.
Businesses are using new technologies to give themselves a competitive edge in the marketplace. They also are employing technological innovations to provide customers with superior or more cost-effective products and services.
Gene Rebeck is TCB’s northern Minnesota correspondent.