85% of CEOs surveyed in PwC’s Global CEO Survey are convinced that AI integrated ERP will have a significant impact on companies and their business models over the next five years.
What is an ERP system?
ERP system is a type of software for organizations to manage daily business activities of various fields such as finance (financial analysis, procurement, accounting), operations (supply chain management and planning, inventory management) and human resource management (workforce planning and management).
It is a business asset that eases tracking of project processes and resources of an interconnected and interdependent business. ERP systems
- connect a wide range of business processes and enable data flow between them.
- collect data from different sources of the business and create a common database
All the organization can rely on the database to reach the up-to-date, correct data since each stakeholder creates, uses and stores the same data derived from the central data archive.
What are the companies in the ERP ecosystem?
The ERP ecosystem has these major parts:
- Software vendors
- Consulting firms or independent consultants
- Adopting organizations: ERP systems are used by businesses of industries and in all sizes for managing various kinds of enterprise resources. Manufacturing companies form a major segments and value industry specific features. A study by Aberdeen Group found that 48% of manufacturing business leaders said that they required such technology to contain sector-specific best practices. Companies of professional and industrial services, healthcare, distribution and construction industries are also a part of ERP adopters. Data from HIMSS Analytics shows that the adoption of hospital ERP software has grown to 38% in 2018 from 19% during the last decade
Which AI technologies are relevant to ERP systems?
A machine learning enabled ERP can benefit companies through
- Predictive analytics (i.e. more accurate forecasts)
- Root cause analysis
- Optimization (e.g. yield rates at the machine, production cell or plant level can increase by optimization of production parameters leading to lower material consumption or higher quality output
Agents with conversational AI capabilities can provide an easier user interface to ERP systems which don’t tend to be user friendly.
What are AI use cases in ERP?
In general, AI models enable organizations to reduce costs and improve operations. Examples are:
Advanced Analytics & Forecasting
Most operations activities such as supply chain management and production can benefit from accurate predictions. AI models can improve predictions using historical data and current conditions. More specific applications exist in:
- Production: Better managing seasonality to avoid underproduction or overproduction.
- Warehouse Management: For example, BMW uses learning algorithms to track an item from the manufacturing stage until the moment it is sold, monitoring 31 assembly lines in different countries.
- Sales: More granular analysis of sales can provide better forecasts which translate to better targets, improving employee performance
Most ERP systems provide basic HR functionality. However, increase analytics capabilities can help companies improve performance management.
IT and Finance/Accounting
In an Harvard Business Review survey, 34%-51% of business leaders predicted that by AI will have its biggest impact in the enterprise on their back-office functions of IT and finance/accounting. Specifically, AI in financial management can
- automate repetitive accounting functions
- increase efficiency of transaction-processing
- verify accuracy of statements and reports
For more, feel free to read our research on AI in finance.
AI integrated ERP can allow quicker, cost-effective, and consistent service. A widely cited benefit is chatbots that are used instantly to answer customers’ common inquiries. Thus customer service representatives can handle more complex customer queries.
For more, feel free to read our research on AI in customer service.
What is the level of interest in integrating AI in ERP systems?
IDC estimated that by 2021, a fifth of the Forbes global 2000 manufacturers will be using IoT, blockchain and cognitive (artificial intelligence and machine learning) to automate large-scale processes. Impact from intelligent ERP was expected to be a critical KPI by 15% of G2000.
However, by 2018, only 16% of industrial companies were using IoT data in ERP and in many companies, ERP was used for Excel runs production. So, even though there is an interest in smarter ERP systems, putting it into practice is taking time.
What are the challenges of using AI technologies in ERP?
Data quality: Machine learning predictions can be as accurate as the data quality of their training data.
Data collection: Most ERP systems serve specialized functions. This leads them to collect only a small portion of enterprise data which limits training data for AI models. According to a survey of 263 IT leaders:
- 94% of companies use less than 80% of the features in their ERP system
- ±33% use less than 30% of the features
What are example case studies of companies using AI in their ERP systems?
As a drug wholesale company based in the US, the company had previously used spreadsheets to pull in data from various systems to determine production costs. After pulling data, historical information and know-how of the employees were also considered figure out how sensitive customers were to price changes.
They then moved to an integrated system that automatically calculates production costs, analyzes historical transaction data, and pulls in outside data such as weather forecasts to create a foundational layer for future deployment of artificial intelligence.
With the old manual system, pricing team members needed to spend 3 hours on more complex price analysis, and 5 hours on the more routine tasks involved with price administration. With smart automation, they are able to spend just 1 hour on price administration, and the other 7 hours on value-added activities.
After implementation of the AI and process automation in Oracle Cloud, company claims to have achieved
• Uptime has increased 60%
• Production has increased by 30%
• Manual processes reduced by 55%
• Floor space has reduced by 85%
Walmart, as an early adopter of HANA by SAP, claims to have been able to process its high volume of transaction records (the company operates more than 11,000 stores) within seconds.
What are the ERP software with AI capabilities?
With increased interest in AI, every major ERP vendor claims to have integrated AI capabilities in their offering. It is impossible to verify all of these claims but here are vendors that claim specific improvements in their ERP solution thanks to machine learning.
- Infor Coleman: Provides conversational UX with chat, voice, and image recognition capabilities
- NetSuite Intelligent Cloud Suite claims better insights and greater efficiency through integration of predictive analytics
- SAP S/4HANA Cloud is a real-time enterprise resource planning suite built on an advanced in-memory platform. HANA enables companies to use machine learning algorithms on their data or build their own solutions with connectors to HANA. predictive analytics, for example, HANA can facilitate cost forecasting to reduce budget overruns and make more accurate resource investment decisions.
- Microsoft Dynamics AI provides virtual agents, sales insights and customer service insights thanks to machine learning.
- SYSPRO claims that its ‘digital citizens’ AI-powered bots are able to integrate directly into the ERP to deal with repetitive tasks across the organization.
- Epicor EVA, AI-based voice command user interface for productivity improvement. It is claimed to streamline routine tasks like quoting and scheduling. It also provides data-driven recommendations (e.g. anomaly detection in manufacturing)
How to evaluate modern ERP systems in line with advances in AI?
Machine learning capabilities are not the most important criteria in ERP selection. Companies should select ERP systems in line with how they will benefit them while running their daily operations. However, the below factors are important to ensure that the ERP system is future proof when it comes to machine learning:
Effective data management: Companies rarely have a chance to modernnize their ERP systems since these are critical production systems that have been deeply integrated into the companies’ operations. So companies need to make sure that when they switch to a new ERP system, it is flexible enough to store and provide company data in granular detail, in line with company’s operations. As long as data is easy-to-access, companies could use the machine learning components of their ERP or other software to build machine learning models to solve their operational problems.
Ease of integration: No single company should be expected to be the company’s machine learning software provider since machine learning impacts every aspect of a company’s operations. Ideal ERP software should be easy to integrate for 3rd party providers.
For any questions on ERP and AI, feel free to reach out to us:
How can we do better?
Your feedback is valuable. We will do our best to improve our work based on it.