HomeData AnalyticsPlanes, computers and books: Three ‘dynamic scheduling’ use cases from Hitachi

Planes, computers and books: Three ‘dynamic scheduling’ use cases from Hitachi

There are three strands to industrial transformation, reckons Hitachi Vantara, the digital change unit of Japanese hardware and software maker Hitachi. The point is to eliminate, or at least manage and reduce, downtime, bottlenecks, and defects.

Greg Kinsey, the firm’s vice president, wants to deal with each these in turn, in reverse order. Defects have been dealt with, under the header ‘predictive quality’, in a separate piece. Downtime will come later. Here, Enterprise IoT Insights discusses with Kinsey how industrial operatives can manage capacity in their production plants.

He puts the question: “What if you could predict and prevent production bottlenecks?” Elsewhere, he explains Hitachi sets a 3:1 return as the first condition for engaging in any kind of transformation work. With ‘bottleneck management’ – or predictive workflow, or predictive capacity – the company has seen returns of 10:1 in some cases, he says.

“It was with an aerospace manufacturer, and was really quite a simple solution; we just digitised the whiteboards.” There are two problems with whiteboards, explains Kinsey: they are static, and they have no memory. “You have to be in the vicinity, and the data gets erased; 336 becomes 213, and no one ever remembers it was 336,” he says.

The use case is explained below; it provided a lightbulb moment for both the client and provider about how to digitise business.

Hitachi hit upon an even more formative use case, also detailed below, at its own site in Ōmika, in Japan. “We developed an algorithm to predict bottlenecks before they happen,” says Kinsey. “We call it ‘dynamic scheduling’.” The bottlenecking solution works on the same big-data principles as the quality one, described in the last article, and the availability one described in the next.

Kinsey explains: “It looks at a number of factors – incoming orders an outgoing materials; the flow through the factory. The more you know, the more you can anticipate change.”

At this point, we hand the story over to Kinsey.

USE CASE 2.1 | PREDICTIVE BOTTLENECKS (PLANES)

“Every factory I visit has whiteboards; every factory uses paper. I see cars going down these highly automated factory lines in German auto plants with paper in their windows. Industrial transformation is not about robots, in the first instance; it’s about people and non-digital data. When I meet a client, the first thing I do is visit the shop floor.

“With this aerospace manufacturer, the hypothesis was to digitise its whiteboards to share data horizontally, and enable analysis of historical data. I can’t remember how much it cost,; but it was clearly less than €500,000, say. The benefit was in the range of €5-€6 million, however – because this company could de-bottleneck its production line, and its whole problem was it couldn’t keep up with orders.

“It also solved a headache for the CDO, trying to figure out how to digitise his company, because it provided the starting-point. It turned into one of our very first digital projects, with a really high ROI. European companies now understand it’s integral to the whole 10-year factory-of-the-future story.”

USE CASE 2.2 | DYNAMIC SCHEDULING (COMPUTERS)

“We make SCADA systems in our factory in Ōmika; these big-box computers, the size of a refrigerator, used for supervisory control in power plants, and other industrial settings. They are made to order. It is a low-volume, high variation game, and the orders always surprised the logic of the production flow. We’d get bottlenecks, our lead times would drag, and our customers would threaten to go elsewhere.

“So we built up data set and an algorithm; it was the same agile process, with our own people in our own factories. The algorithm revealed the bottlenecks, and proposed how to re-sequence the orders and move people around, or when to switch over machines – this kind of dynamic scheduling of the production flow. As a result, we removed bottlenecks, and cut the lead time in half.

“The Ōmika factory served as test-bed for us. We were able to build out our model for dynamic scheduling, which has since been presented around Europe, and been received with welcome arms. Manufacturers see it as a silver bullet.”

USE CASE 2.3 | DYNAMIC SCHEDULING (BOOKS)

“Book printing, or book manufacturing if you like, is a good example of mass customisation. You can print thousands, and each one is unique. Demand is predictable on one hand, and unpredictable on other – the printing process is highly demand responsive. People order online, and want the book next day, not next week. It’s a different industry, but it’s the same problem.

“We ran the same dynamic scheduling model on a book factory in Japan, with the same result: higher throughput, lower lead-times. The big takeaway, from these three, is the use cases transcend vertical markets. And that’s new – because in the old Industry 3.0 world, you had one application for the food industry, and another one for the automotive sector, and every car company used the same app.

“In this new world, the logic goes transcends the market, and the level of customisation is turned right up.”

This is an excerpt from a wide-ranging interview with Hitachi Vantara vice president Greg Kinsey. For more from the interview, check out the links below.

The IIoT interview (pt1): “It’s a two-speed market; the US doesn’t get it,” says Hitachi

The IIoT interview (pt2): “Three is the magic number for digital ROI,” says Hitachi

The IIoT interview (pt3) “We’re selling innovations, not solutions,” says Hitachi

Drugs, steel and tyres: Three ‘predictive quality’ use cases from Hitachi

Motor cars and gas turbines: Two ‘predictive downtime’ use cases from Hitachi

 

Previous post
IoT America launches managed IoT service for rural US
Next post
The IIoT intervew: "Edge computing is the death of the cloud as we know it," says IIC