“For hundreds of years, humans and machines have worked together. But something peculiar has happened recently: the machines are now speaking to us—and to one another”
-- Booz Allen, Vice President at Sedar LaBarre [1]
That would be a great quote from Booz Allen, if it weren’t slightly ahead of its time. New machines want to speak to us and want to be part of orchestrated processes in the factory, but we are not there yet. The mission ahead for Smart Manufacturing initiatives is to make it happen and realize the enormous potential from new levels of synchronization spanning from product design to order, realization, and delivery.
How Smart Manufacturing complements Lean Manufacturing efforts
Smart Manufacturing can be viewed as a process improvement movement complementary to the Lean Manufacturing movement of the past two decades. Lean Manufacturing initiatives design out the overburden, inconsistencies and waste to achieve enhanced processes that deliver products smoothly and consistently. In contrast, Smart Manufacturing initiatives are bringing the physical and digital worlds together designing in connectivity and orchestration to achieve enhanced processes that deliver products and their data smoothly and consistently to customers. Going forward, both Lean and Smart Manufacturing will be interrelated improvement efforts in the manufacturing value chain.
Eliminating the Hateful Eight Types of Disconnectivity
There are eight types of disconnectivity in the manufacturing value chain that need to be eliminated as organizations embark on the Smart Manufacturing journey to synchronize, automate and optimize the physical and digital processes.
Each of the communication and process participants listed in this slideshow need to be systematically connected to achieve the desired Smart Manufacturing goals, which we discussed in the prior article titled: “On the Journey to a Smart Manufacturing Revolution.” [2]
#1 Equipment
Machines and equipment are getting smarter with more onboard computer processing power and more connectivity to the Internet. This advancement in connectivity is referred to as the Internet of Things (IoT). For example, at home, I can set up my printer to directly order ink when it is low and have it shipped to my house. At the factory, we refer to this type of internet connectivity direct from machines as the Industrial Internet of Things (IIoT).
Equipment, machines, robots and advanced sensors can be connected to systems that automatically adjust or switch operations based on sensing the product, diagnostic or environmental conditions. IIoT connectivity needs to be orchestrated in the overall Smart Manufacturing architecture to, for example, avoid excessive downtime at the factory, because each of our machines has auto-requested maintenance a few days apart instead of coordinating a single service day of downtime for the entire line.
According to a research report from the analyst firm Berg Insight, the installed base of wireless IIoT devices reached 10 million in 2014 and will grow at an annual growth rate of 27% to reach over 43 million by 2020. [3] Yet, many organizations are struggling to reap efficiencies and insights from the connectivity and data because they haven’t created the proper technical and decision making structures to organize connected processes and analyze data for real business value.
#2 Workforce
Yes, people are an integral part of Smart Manufacturing.
People are resources that need to be connected to the processes orchestrated and data produced by the Smart Manufacturing system. We might appreciate disconnectivty when we are on vacation, but definitely not at work. We want to know what is happening at our operations as soon as possible. We want to be able to see when customers raise issues, when the production line is not running smoothly, and when problems are not being resolved in a timely manner.
Technicians, inspectors and material movers need function-specific applications that dispatch them to prioritized tasks and enable them to update status and document issues immediately. Some functions are performed at a desktop, other functions are performed on-the-go on a mobile device—function specific applications connect people at the place where they need to perform the function.
Some workforce resources might be external to the organization. As an example, smaller manufacturers can’t afford to keep every specialized skill as a full-time employee, and might need to route tasks to third-party contract technicians.
Managers need dashboards that display status, trends and alerts to focus their attention on where it is needed most. However, the old passive information model of aggregating data and providing metrics to managers to digest and take action is not enough to achieve new levels of efficiency.
New active manufacturing systems need to automate routine decisions and trigger action based exceptions and unpredictable patterns that require experts to manually analyze for root-cause. There are always plenty of exceptions to manage in daily operations, so don’t worry, people will always be needed.
Legacy paper-based procedures promoted departmental silos by creating paperwork between people in different disciplines, departments and companies in the manufacturing value chain. Social communication and collaborative work tools can connect teams and communities in new ways breaking down old silo boundaries and creating new levels of efficiency in processing and trouble shooting procedures.
As the manufacturing enterprise becomes more connected, it creates the need for a new breed of talent—one that embraces computer skills as much as mechanical skills. The next generation of manufacturing talent must be able to collaborate across multidisciplinary walls, connect digital factories, interpret results from analytics, and protect supply chains from cyber intrusions. Luckily, millennials are tech savvy, learn fast and embrace teamwork. Manufacturers will face strong competition from high-tech companies in recruiting this talent and will need to elevate the importance of strategies to recruit, develop and retain the workforce of tomorrow.
#3 Design
There are currently many manual steps and translations between the virtual product and process designs modeled in 3D and the programming and configuration of physical manufacturing and inspection processes. This type of connectivity is referred to as the “digital thread” spanning from design into supply chain, manufacturing, inspection, and product service.
One of the first applications of the digital thread is between product design and 3D printing processes. A few more of the enhanced processes targeted for the digital thread include: (a) simulation of the virtual production processes and validation of designs in collaboration with partners before physical processes are ever changed, (b) flow of revised specifications into the supply chain to ensure subassemblies are received to the correct design revision level, and (c) automation of translations between design specifications and instructions loaded into automated production and inspection machines.
#4 Parts
One of the goals of Smart Manufacturing is to create flexible manufacturing processes that can adapt to build different products and support mass customization business models. Connected auto-identified parts and materials can help achieve that type of flexibility.
Accuracy of product genealogy becomes more difficult to achieve as product variances and changes increase in a multi-tier supply chain. Accuracy can only be guaranteed if identifiers are read directly from components and materials with automatic identification (auto-ID) tags, like RFID, either embedded in the component or in the packaging. Common applications for RFID include tracking goods in the supply chain, reusable containers, high-value tools assets, parts on inventory shelves, and parts issued to manufacturing.
Flexible manufacturing processes need to (a) recognize the configuration of each product as it traverses the line, (b) load the right parts and programs to machines, and (c) compare component parts as they are loaded into machines and assembled into product in order to verify the correct bill of material configuration.
RFID tags can contain more than part and serial numbers, they can have information about hazardous warnings, expiration date or storage requirements that could trigger special handling processes at the shop floor. For example, automated processes could alert staff if material has expired due to the time it spent outside of the required refrigeration temperature.
#5 Processes
Business processes are chains of tasks that produce a specific service or product for specific customers. Tasks in a business process can span across organizational departments and system boundaries. In order to minimize delays throughout the business process, task outputs need to be connected as inputs to successor tasks. Communication and data processing among tasks should avoid manual data input and translation errors whenever possible.
Legacy business processes where designed around the sequential handover of paper documents between different departments. Modern business processes can be reinvented around new cyber-physical paradigms that promote real-time response, collaborative teams and parallel tasks across production and supply chain. Consider the benefits of processes where utilities auto-adjust based on environmental sensor data, where machines take corrective action to avoid damages, where part shelves report usage and are automatically replenished by suppliers, where correction tasks for non-conformances are routed in parallel to multiple departments including Engineering, Procurement and Inventory Control.
Automated processes are standardized processes, and standardization promotes best practices and reduces overhead tied to maintenance of multiple ways of doing similar functions in the organization.
#6 Suppliers
To optimize today’s highly outsourced manufacturing value chain, supply chain management has to switch off the traditional procurement function managed by paper documents, emails and phone calls. Suppliers need to connect digitally with the companies they supply via web portals, supply chain hubs, and business-to-business (B2B) communication exchanges.
A demand driven supply chain can reduce inventories and schedule performance. Automated demand triggers can be achieved to replenishment and move material based on sensing part installation or empty shelves.
The orchestration of multi-tier supply chains for complex products like aircraft or medical devices require additional attention to connected change management processes for product design information.
#7 Customers
Manufacturers have traditionally faced challenges in collecting reliable data to understand and adjust to changing customer needs. Data collected via surveys is unreliable, expensive and burdens the customer.
Today’s data analytics capabilities represent a new frontier for customer satisfaction. The ability to collect and analyze customer usage data in real time allows manufacturers to make quicker adjustments. For example, remotely collecting usage data on farms from deployed windmills can reveal shifting wind conditions and immediately allow a remote operator to optimize the degree of effort that specific windmills are exerting within those farms. This type of data visibility direct from customers allows manufacturers to not only fix current products, but also anticipate the need for new features before customers ask for them.
Customer involvement in the middle of some business processes can lead to better customer experiences. Customers want more visibility of progress on long projects, and online processes to coordinate orders, deliveries, audits, warranty claims and returns. Customers should be able to remotely approve changes like deviations from specifications, late deliveries, and corrective actions.
#8 Records
As more machines are connected, it would be easy to fall into data hoarding practices. Data should be stored for a specific period of time as part of a connected system strategy, and not just collected because it might be useful one day.
Data hoarding practices can be expensive and make it hard to find real productivity improvement areas among all the noise.
Different documentation and data records have varying degrees of precision, control and time persistence requirements in support of different business processes in the organization. Some data can be of a temporary nature and only needed during the execution of a specific business process, but some data has persistence requirements.
It is important that systems of record are well connected as sources and destinations for data in processes orchestrated by Smart Manufacturing systems. Regulated industries require detailed production history records. As an example, the FDA's quality system regulation (21 CFR Part 820) requires medical device manufacturers to establish and maintain device history records (DHRs) for each batch, lot and unit they produce. The DHR must include as-built history that can be validated against the respective product design (as-designed) and applicable processing and labeling standards.
Pulling it Together With Security and Communication Standards
Security and communication standards are two important considerations that apply to each of the eight areas of connectivity for smart manufacturing systems.
The connected nature of future Smart Manufacturing systems provide great opportunities for optimization but also increase the exposure to potential cyber threats. A strong cybersecurity program spanning people, processes and technologies is necessary to protect information assets and the trusted relationships with suppliers and customers.
A key missing capability to achieve Smart Manufacturing is the ability to practically and affordably knit together a threaded orchestrated business processes across varied equipment, systems, suppliers and customers.
Today, vendors provide custom one-off solutions that are complicated, expensive, require enormous amounts of time to implement, and are built on proprietary communications. To achieve affordable plug-and-play capabilities, next-generation hardware and software technologies need to work together through common standardized communications.
The list of disconnectivity areas above serves as a roadmap for identifying gaps and opportunities towards new levels of connectivity for future Smart Manufacturing systems. They are all of equal importance on the journey to future Smart Manufacturing excellence.
References
[1] “Capitalizing on Connectivity: The Top 5 Priorities for Industrial Manufacturers”, Booz Allen Hamilton Inc., January 2015
[2] “On the Journey to a Smart Manufacturing Revolution”, Conrad Leiva, IndustryWeek IdeaExchange, December 2015.
[3] “The installed base of wireless IoT devices in Industrial Automation reached 10.3 million in 2014”, Berg Insight blog post, May 2015
[4] “5 Signs you are a Big Data Hoarder”, Bernard Marr, Data-Informed.com, August 2015
[5] “MESA White Paper #52: Smart Manufacturing – The Landscape Explained”, Smart Manufacturing Working Group, MESA.org, January 2016
This article was originally published as "Eliminating the Eight Types of Disconnectivity in Smart Manufacturing" at IndustryWeek.com, Feb 2016
Comments