Intelligent supply chain is the reinvention of the supply chain with emergent and intelligent technologies that have matured to practical use:
- The Internet of Things (IoT) makes business applications interact with the physical world
- Big Data makes large data sets accessible for advanced analytics and intelligence
- Machine learning (ML) and artificial intelligence (AI) automate repetitive processes and learn from human exception handling and decision-making
- Advanced analytics finds data patterns to support decisions and predict the future
- Blockchain distributes collaborative processes across the entire value network
- Data intelligence finds new value in data assets for new business models
Intelligent technologies like these are poised to automate and optimize the entire supply chain, from design, to manufacturing, to asset management, to logistics, to after-sale services, all with the purpose of delivering customer value in products, services, and brand experience in an agile and responsive fashion.
Defining the intelligent supply chain
An intelligent supply chain connects people with processes and things to enable visibility, communication, planning, analysis, simulations, and execution. These processes are orchestrated across the different functions based on real-time inputs and requirements coming from connected assets and products through IoT infrastructure.
Intelligent supply chain is the transformation of a conventional supply chain into a digital supply chain by re-imagining the various processes and infusing them with intelligent technologies. It connects all the physical assets that make up the supply chain to the digital model to extend visibility, control, and simulation of operations to optimize decision-making under uncertainty.
One fundamental aspect of an intelligent supply chain is the continuous capture of relevant data generated through the manufacture, transport, and other assets that touch those products and services. The various streams of data are aggregated and combined into a digital model of the physical supply chain. The application of AI and ML to this model creates insights to drive intelligent decisions.
Another fundamental aspect is the flexibility to react quickly to changing customer priorities. The intelligent insights captured by the digital supply chain model help to anticipate potential issues or opportunities and capitalize on them. Increasing agility by being able to identify bottlenecks and quickly simulate real-time alternatives enables organizations to respond to potential problems in an optimal way.
Where intelligence is applied today
Solutions providers have been embedding intelligent technologies in their applications to address customers’ problems. The aim is to seamlessly bring together cloud platforms with Big Data, advanced analytics, machine learning, and IoT. Connecting this ecosystem to a core enterprise resource planning (ERP) solution and combining operational and customer experience data provides insights that support better decision making.
Some of its main applications today can be found in the following areas.
Intelligent asset management
IAM provides a digital representation of all industrial equipment with secure, controllable sharing rights across business partners to enable customer insights, performance analytics, and new digital services. The outcomes sought include maximization of asset productivity, safe operations, and cost reduction.
It covers the entire asset lifecycle: engineering, procurement, construction, commissioning, operations, inspections, overhauling, upgrades, and decommissioning.
Companies are building digital twins of their assets to monitor them throughout their lifecycle in a closed loop, where engineering receives feedback about operations from customer sites. This enables detection of problems before they become critical. Equipment-as-a-service, remote diagnostics, and predictive maintenance are possible based on the continuous stream of data from equipment and analysis by AI algorithms.
Predictive maintenance
IAM is also enabling predictive maintenance to become a reality.
Machine learning coupled with digital twin models can detect patterns of malfunctions through prediction and anomaly detection algorithms before their effects become evident in the equipment performance. This provides a longer lead time for predictive and prescriptive measures to avoid costly downtime.
With the right technology, asset data collected by IoT sensors and gathered through the cloud can be analyzed in real-time so companies can monitor equipment usage and proactively advise or recommend service or maintenance activities.
Service agreements and remote troubleshooting can be improved through predictive analytics, machine learning, and IoT-fed virtual digital twins of deployed assets that can model product performance and diagnose (or even anticipate) hardware or software problems.
In addition, advanced augmented reality can improve maintenance activities by overlaying drawings and schematics on a machine, visualizing sensor information, and receiving repair instructions over mobile devices (such as smart glasses); these are becoming a reality in many manufacturing companies.
Connected fleets for digital logistics
Connected vehicles enable companies to adjust in-route shipments and optimize delivery routes based on real-time weather and traffic conditions to maintain delivery dates or minimize the impact of unexpected events. This technology also allows real-time tracking and monitoring of conditions, such as freezer container temperatures, to guarantee product quality.
This model increases productivity, gets goods to customers faster, and improves fleet transparency by collecting, mapping, storing, and analyzing telematics and sensor data.
Global track and trace
Sensor networks in products and transportation fleets are enabling companies to sense and respond to logistics insights, providing real-time insights to relevant stakeholders for their business and compliance needs. Business events are captured across the extended business network environment by connecting them to the IoT infrastructure, ensuring tracking of processes, materials, products, and assets with scalable performance and throughput.
Warehouse automation
Intelligent supply chain management is changing the game for traditional warehouses, retailers, consumers, and employees. The use of augmented reality (AR) for picking and packing processes is streamlining and optimizing warehousing activities, increasing efficiency, and reducing time from shelf to delivery. Together with automation using robots for transfer operations, these technologies are enabling companies to shorten their delivery time and keep their promises to customers.
Integrating automatic guided vehicles (AGVs) in warehouses with ERP and manufacturing execution system (MES) software, manufacturers can better manage inventory levels and improve on-time delivery. This ensures material is delivered to production and between operations at the right time to ensure minimal downtime. This enables the 24-hour automated factory of the future.
Manufacturing planning and scheduling
Industrial manufacturers are using machine learning algorithms, a key component of artificial intelligence, to leverage available, valuable data sources and employing Big Data technologies to improve manufacturing planning and scheduling processes.
Smart sensors that provide critical status data can assist with the automatic routing of products to work centers in the production process, increasing efficiency in asset use and providing real-time feedback to schedulers and planners. This cycle improves the accuracy of planning and the real-time simulation of alternatives when problems occur on the production line.
Pay-per-use business models
Companies, especially those in the industrial manufacturing and components sector, have started to offer pay-per-use as an additional business model for software and digital services, as well as for machines and equipment. Equipment-as-a-service contracts for auxiliary and core business products will become commonplace in the near future.
Manufacturing companies continually review their strategies to drive profitable growth in difficult markets. To compete, large manufacturers are evolving from selling capital-intensive machines to providing equipment to customers and charging by the number of successful jobs completed. This operating expenses-based billing is very popular with customers, as risk is shared.
Due diligence with machine learning can recommend low-risk customers for this pay-for-outcome model. Pricing can be automatically generated based on customer risk factors and typical costs incurred in discrete manufacturing, such as installation, transportation, consumables, etc. The technology solution learns with every new contract and corrects its mistakes. Discrete manufacturers can focus on partnering with their customers, not just providing machines. It’s a first step toward running a successful intelligent supply chain.
Blockchain
Blockchain is being used in the supply chain in many ways, including:
- Asset management (digital twins)
- Distributed manufacturing (3D printing)
- Trusted digital credentials
- Secure bidding in procurement
- Drug supply chain security
- Asset tracking
- Tracking of mandatory inspections and audit information
- Pay-per-use scenarios based on sensor data and micro-payments
- Parts provenance
- Equipment transfer of ownership
Digital twins and IoT as the backbone for the intelligent supply chain
From the perspectives of manufacturers and asset operators, digital twins in the supply chain give visibility into products and equipment as they are designed, configured, procured, built, installed, and operated. They collecting real-time data throughout a product or service’s lifecycle, closing the loop from operations to engineering.
Understanding how products are used and perform in real-life conditions, once they are in customers’ hands, is crucial for making informed engineering decisions on product upgrades, augmentations, or add-ons.
IoT and embedded sensors provide continuous feedback on products’ performance in the field. Data can be collected in the background and in real-time without interfering with usage. This, in turn, can directly inform new products, features, or upgrades that better meet customer requirements.
Challenges ahead
As with any disruptive technology, there is no clear path nor best practices to lean on. Solution providers are building those practices and pathways based on experiences with pilot and implementation projects. The situation is exacerbated by the fact that several emerging technologies have been converging in recent years as the foundation for Industry 4.0. There are now many different names for essentially the same thing: intelligent enterprise, intelligent supply chain, smart factories, factories of the future, etc.
The newness of the technology and the fact that so many vendors are trying to establish a footprint in the industry make it very difficult for customers to decide which platform to use or which strategic direction to take (even though all agree that doing nothing is a bad strategy). That is one reason why so many projects don’t move from proof of concept or pilot implementation to enterprise-wide deployment. There is too much confusion, and the speed of technology maturity is much faster than a company’s leaders can absorb and assimilate to make strategic decisions related to the intelligent supply chain.
Since these technologies are relatively new and evolving, leading companies have been piloting different solutions as a way to learn and build a business case for deployment at scale. Unfortunately, many haven’t been able to exit the pilot phase and scale it to operations and the business due to the reasons cited above. We have seen success in small-scale projects, such as one warehouse in one manufacturing plant; garbage collection in a city; selling compressed air-as-a-service instead of selling air compressors; and embedding sensors in electrical motors to sell uptime-as-a-service and provide remote maintenance.
As in all past disruptions, the path is not totally clear, but the fog will lift as quickly as we move forward, learn from our mistakes, develop the necessary capabilities, and ultimately transform processes and business models to create the intelligent enterprise.
Data is the new oil. SAP Leonardo Hierarchy Matching serves as a ‘refinery plant’ for your master data by using Machine Learning capabilities to solve various business problems. Find out more in our webinar on June 11.