How embedded business AI facilitates the next wave of automation and transformation.
In light of the omnipresent debate about the role and potential of Artificial Intelligence (AI) as disruptive force to global business organisations and existing processes, this blog intends to provide an assessment of SAP’s approach to business AI and the application potential for complex manufacturing organisations, such as in the Aerospace and Defense (A&D), or other Engineer-to-Order (ETO) industry sectors.
AI Impact on the A&D Industry
With breakthroughs in traditional and specifically generative AI, A&D / Manufacturing executives have increased their focus on evaluating AI-related initiatives, in disciplines like:
- robotic process automation (RPA)
- traditionally applied ML e.g., to predict outcomes and probabilities or to discover correlations, anomalies, or clusters, hidden inside massive data sets.
- deep learning / Gen-AI applications focusing on natural language processing or image and video object detection and content creation.
There is consensus among industry experts that AI opportunities exist almost everywhere – from core product R&D and air mobility innovations, over optimisation of flight operations and air traffic management to enabling new business models, service, or cybersecurity capabilities.
The focus of this blog will be on opportunities in the core business process framework for commercial and defense aerospace and defense enterprises as well as other complex, ETO centric manufacturing organisations.
Seeking effective ways to respond to current industry challenges and market realities impacting all of A&D and adjacent manufacturing sectors, many companies assess the potential of AI solutions starting with production, supply chain or maintenance operations, hoping to gain much-needed improvements in operational performance or quality metrics. Whether these aim to improve “first time right” scores, order throughput rates or getting more accurate predictions on supplier performance or product margins, many look at AI as a must-have tool for survival in a challenging environment.
The SAP approach to Business AI
SAP’s approach to Business AI, is based on three core principles: It has to be relevant, reliable, and responsible.
SAP’s existing business process platform has been providing best practices and end-to-end solutions for mission-critical industry processes for several decades. By natively embedding AI into this rich foundation of business process and industry data models, we make AI innovations available to enterprises directly where they have the biggest impact, tailored to their specific business context and allowing to re-think existing processes.
Interested in more background? Please start here or take a few minutes to review this blog for a more detailed review of recent examples of selected AI innovations.
Figure 1 - Three sources of value unlocked with SAP Business AI
Business AI applied: A journey towards a future-proof digital enterprise.
With an embedded business AI approach, the AI can work directly on the proven industry data models that SAP has built with thousands of customers over decades and that it can train on the semantics and correlations across your unique custom enterprise data model and landscape – of course considering data protection, ethics, and data privacy standards.
To learn more about SAP’S AI ethics policy, please take a look here.
The enterprise data universe on your SAP platform already collects significant knowledge about actual operational performance values such as supply lead times, non-conformances information, as-built genealogy, or product cost details, that would allow AI services to draw detailed conclusions and arrive at more accurate and relevant predictions on performance, quality, or risk metrics – tailored to each companies’ context.
While we see a broad set of common use cases and value drivers across the industry, we also expect companies to approach AI at their own pace, with their own priorities, and in context of their own realities.
The following structure should therefore only help to exemplify the key patterns and examples that emerged from initial customer and internal conversations.
Figure 2 - Vision: How AI could transform A&D in the future
For most corporations, this journey does not start today or tomorrow, but driven by urgent needs to improve operational efficiencies and master the overarching workforce challenges, companies have already started to look into applying proven technologies that we summarise here under Horizon – 1: Automation and Optimization.
This covers existing AI technology to increase automation of repetitive tasks and processes with robotic process automation, intelligent workflows, and optimize processes with situation handling, or ML processing of unstructured information.
In the same category we also include solutions to provide system users with a more natural user experience, such as chatbots, digital assistants or intelligent content summaries. At this stage, Business AI is applied to augment and improve human decision making by providing intelligent business recommendations with insights.
In initial discussions with some of our A&D customers and partners, we have identified a couple of use case ideas, for example:
- Chatbots to assist non-expert users with context-specific system guidance for complex, time-critical processes, e.g., how to assess a production order change scenario or determine program management status information quickly.
- Intelligent cross-system workflows and “lights-out” automation of processes like financial closing and billing preparation to avoid manual errors and speed up processes, drive standardisation and scalability.
- Automate collection and validation of critical master data to improve quality and accuracy, e.g., supplier lead times, raw material prices, or estimated production scrap rates. Getting these parameters “right”, has massive business impact.
Horizon 2 – Intra-company transformation:
True transformation potential lies in the ability to re-engineer entire E2E processes holistically across silos and teams, instead of focusing only on isolated process steps – and if needed by combining multiple AI technologies.
For example, applying ML to forecast equipment and machine failures as well as ideal maintenance schedules based on historical records, and additional context data can help optimize the maintenance cycles for critical assets and increase their uptime.
But what if companies could automatically use these predictions as input into adjacent enterprise processes, such as production planning (e.g., if my AI can estimate that a production asset is ideally taken out for service or overhaul at a specific time in the near future – so a well-known predictive maintenance use case - what is the impact of this capacity change to the already planned production schedule and what is the optimal solution to adjust this plan – considering customer priorities and shop-floor resource and tool dependencies).
A similar example, that some companies have already started to look into: Use ML to predict short-term risks that deliveries of critical supplier parts to a factory / assembly line will be delayed due to external factors disrupting transportation routes – combined with a recommendation engine that analyses the impact on the production flow and the best options how to re-schedule and move orders in different scenarios.
A Business AI solution that understands all these connection points hidden in the business data model would have significant potential to help companies optimize entire process chains, not just individual sub-processes, for maximum impact.
Horizon 3 – AI on network level cross-company
Of course, the long-term vision of business AI goes far beyond intra-company optimization but could lie in a much more autonomous design and implementation of AI-based processes reaching across company boundaries and connecting information and processes from multiple networks and system layers and hereby moving to a future state in which systems could take more autonomous decisions with high business impact – enabled by intelligent bots that optimize cross-network business flows, e.g. in supply chain planning and fulfillment and also provide more transparency for supply chain planners.
How to apply AI use cases to drive innovation across the A&D Value chain (some ideas)?
SAP and our global partner ecosystem in the A&D and broader Manufacturing space have already started to work on concrete use cases and innovations across the entire range of business processes that impact the main strategic priorities for companies in the A&D space. Some are already available today; other examples are innovation ideas that SAP or our partners might look into in the future.
To stay up to date, I’d recommend to frequently review the latest SAP AI roadmap information (link provided at the end of this blog).
Customer Centricity (all process related to end-customer engagement, proposal management, and business acquisition)
- Generate more profitable proposals and quotes, recommend ideal solution and service proposals, and optimize re-use of historic projects even for complex ETO environments
- A first step in this direction is our SAP Intelligent Product Recommendation offer for CTO processes. Here is an overview: SAP Intelligent Product Recommendation Spotlight Video
Figure 3 - SAP Intelligent Product Recommendation
Agile Innovation and Manufacturing
- AI-supported quality assurance - Improve production efficiency, product quality and non-conformance control with Image / photo-based inspections, ML-powered quality and yield prediction, failure classification, automated root-cause analysis, and recommendations.
To get a first impression how this could look like in our SAP Digital Manufacturing Cloud solution, take a look at the Visual Inspection scenario.
- Asset utilization benefits from maintenance prediction and health monitoring.
- Worker efficiency - Use AR to provide hands-free assistance, digitise knowledge transfer and recommend learnings and training
Responsive Supply Networks
- AI can help making supply chains more resilient by improving demand forecasts and identifying potential supply chain risks and delivery disruptions early on.
- Intelligent optimization of warehouse tasks and space allocation to minimize storage movements and handling cost.
- AI-supported lead time determination - Predict lead time variances more accurately, to improve planning quality and delivery performance
- Forecasting of long-term service part demand.
New Business Models / Aftermarket
- Predict impact of service parts inventory constraints on SLA, use ML to allow planners to run scenarios of potential risks and outcomes to assess impact on performance contracts
- Support authoring of maintenance engineering and planning documents, detect failure patterns in unstructured maintenance records
- Predictive asset maintenance - Leverage IoT and other data sources to proactively identify maintenance issues and use ML for failure modes, anomaly, and root cause detection. Interested how SAP Asset Performance Management works? Take a look here.
Figure 4 - SAP Asset Performance Management
AI - What is next …
With this blog, I have tried to summarise some of the main application and use cases ideas along the typical design-to-operate process flow for manufacturing companies, specifically in the A&D sector.
In one of the next blogs, I’m going to look into some additional aspects in the areas of defining a governance structure for successful AI initiatives and share some updates on innovations.
In the meantime, feel free to reach out to me if you are interested and want more information about use cases and scenarios for Business AI for the A&D and Complex / ETO Industrial Manufacturing Industry sector.
You can start your own research, here are some additional sources:
- SAP AI Roadmap with focus on A&D
- SAP Store - Partner Solutions for A&D Industry, powered with AI.
- Explore SAP for A&D
- What is generative AI?
Thank you / Best Regards,
Frank Klipphahn - SAP Industry Expert: Industrial Manufacturing and A&D
- SAP Managed Tags:
