Building a Smarter Control Tower:An IBM Supply Chain case study

Known for its software, cloud and services offerings, Big Blue also is a leader with its USD 7 billion hardware and systems business.¹

The supply chain that helps sustain the company’s market leadership is complex, bringing together hundreds of professionals executing more than USD 40 billion in materials and services spend on behalf of its own business and as a service for clients.

The company prides itself on operating a first-class supply chain organization and has long been recognized for manufacturing leadership, innovation and Supply Chain Social Responsibility.² However, like most best-in-class supply chain organizations, IBM strives for continuous improvement.

• Strategic business risks: Changes in the market and business environment that necessitate adaptation in strategy and product delivery.
• Major supply chain risks: Unforeseen risks and disruptions, including Black Swan events, that have a significant impact on supply chain operations.
• Operational disruptions and risks: Daily and weekly disruptions in the availability and delivery of supply that impact customer orders.

The team knew their biggest challenge was disparate and siloed supply chain data that limited visibility into disruptions and created a drag on response times. But they believed that transformative technologies – including cloud, blockchain, the Internet of Things (IoT) and Artificial Intelligence (AI) – were mature enough to help them address the challenge. The concept of a modern control tower with AI seemed especially promising in helping to drive significantly greater supply chain visibility, forecasting and predictive capabilities to achieve a state of supply chain orchestration.

IBM launched a strategic plan, called the Transparent Supply Chain Initiative, to test the effectiveness of a control tower solution infused with these technologies in addressing its most relevant use cases. The primary objective of the initiative was to bring together supply chain professionals with the intelligence and capabilities needed to better prevent and more rapidly respond to disruptions and events.

IBM supply chain leaders were confident that the Transparent Supply Chain Initiative could enable new ways of optimizing supply chain operations for efficiency, responsiveness and cost – and drive improved outcomes in areas such as supply assurance and inventory optimization.

To troubleshoot inbound supply disruption, the company’s supply chain team had historically used spreadsheets and assembled personnel from around the world via conference calls. A tremendous effort was typically exerted to:

• Identify and prioritize potential disruptions and risks.
• Draw up resolution options.
• Assess collateral-damage impacts.
• Conduct trade-off analysis across the supply chain.

As one IBM executive noted, "We didn’t identify risks and issues early enough. In retrospect, the signs of a disruption were usually evident, but the company did not have enough data and insight to consistently look for them ahead of time. The pace of information and team response went along like a raft on a river. The organization was not consistently looking ahead or steering the raft. It typically was a reaction to events." ³

Leadership recognized that this exercise happened daily in the organization, just at differing scale, speed and impact.

To operate as a single, agile team, leadership knew they needed to be on the same page about the facts of every supply chain disruption, which would be very well addressed with a control tower with near real-time, end-to-end visibility. The team believed it was critical to establish a single source of the truth across the company’s more than a dozen Enterprise Resource Planning (ERP) systems, as well as other internal and external sources of data, including third-party logistics providers (3PL).⁴

As frequent early adopters of new technologies, the IBM supply chain team recognized that AI could help address multiple aspects of the inbound supply challenge. AI could draw data directly from the organization’s ERP systems and other relevant sources, analyze it in real-time, and serve up actionable insights to help the team make better decisions faster and take more confident action to resolve disruptions. ⁴

An AI-enabled control tower could also aggregate Key Performance Indicators (KPIs) and produce related smart-alerts for specific personnel.

As noted in, The Evolution of the Transparent and Cognitive Supply Chain, "Some of the company’s risk management systems provided alerts. However, like an alert on a car dash panel, they didn’t provide information about the nature of the problem, or the technical details on how the problem could be solved. IBM thought it was critical to identify the elements associated with events, as well as the touchpoints in the network that needed to come together to address these issues. The leadership team defined an ideal alert … as characterized by immediate notification accompanied by the right information to make a decision." ⁴

The leadership team also decided to help speed the disruption-resolution process by leveraging virtual collaboration technology to quickly bring together the right being made by cross-enterprise and cross-functional teams. The team named these online collaboration platforms, Resolution Rooms. A new Resolution Room would be established for each key event, and AI would aid in the collaboration and decision-making process by providing intelligence relative to the situation in advance. Natural language processing capabilities allowed Resolution Room participants to ask IBM Watson questions and get immediate answers.

Finally, the leadership team wanted a better way to capture institutional and local knowledge and share it with the global team. They leveraged AI for this task as well. Because AI learns over time – and with each new data point and decision – they knew it was the best way to preserve the organization’s supply chain history, preferences and best practices to build a body of knowledge for the future.

IBM used this capability to develop Digital Playbooks that served as an AI data source and reference when teams were dealing with specific events in a Resolution Room. This enabled exponentially greater speed and accuracy in responding recurring supply chain challenges over time.

IBM’s supply chain leaders know that improvements in inventory management almost always translate into reduced disruptions, greater customer satisfaction, and a direct impact on the bottom line.

Although inventory management systems go a long way in helping organizations manage challenges – efficiencies can be gained by using AI to make better connections and correlations from the inventory-side of the business to the demand-side, including ERP and related supply chain systems.

IBM wanted a closer connection between demand forecasting and inventory management – both organizationally and technologically – to better plan, predict and meet customer demand. Improved inventory insights would enable the company to better and more rapidly adapt to market changes and growth demands.

The IBM supply chain team designed a supply chain control tower solution that leveraged AI to extend and connect existing inventory solutions and ERP systems to more effectively match supply volume to customer demand, and lower safety stock and inventory levels.

The AI-enabled solution helped the team more accurately predict demand spikes and changes – and compare those with inventory – by correlating all structured inventory data across the supply chain and supplementing it with third-party data.

By capitalizing on AI to streamline data integration and correlation, IBM was able to achieve comprehensive visibility and attain:

• More accurate demand and inventory data, analyses and predictions.
• Deeper inventory insights.
• Increased speed and agility in responding to events and opportunities.

The ability to use AI to access, correlate and understand "dark" data was particularly important to the demand side of the business. More than 80% of all data is dark and unstructured – everything from market analyses to news, weather and social media feeds.⁵ Visibility into and understanding of dark data provided additional context for more relevant insights and demand-supply intelligence.

Demand forecasting is a difficult task and applying AI to complement existing solutions with predictions and analyses is a tremendous strategic advantage. As the supply chain team’s AI solution learns more about the internal and external factors that influence IBM’s inventory levels, even greater insights, intelligence and value will be derived from its predictive capabilities.

Ultimately, an AI-enabled control tower allows the supply chain team to more accurately forecast demand and inventory levels and anticipate possible disruptions to determine the best course of remediation, including identifying substitute suppliers and inventory sources, as well as alternative delivery options.