Understanding the history and development of AI can help us determine how we might best apply it to our business practices today. So how do we use AI in Resource Planning?
AI is very good at problem solving and reasoning, once parameters are given to it to. The idea is to build up a set of rules for AI to follow then present the date to AI following a system it understands to sort out resources and best plan them out.
Since the 1950s, we have been fascinated with the idea of Artificial Intelligence (AI). Nowadays, we are just as intrigued with what AI can do for us and what it means for our workplaces and our communities. Discussion about what has been achieved and what is possible is happening everywhere.
AI is already ubiquitous in our world. It is used in translation tools, Satellite navigation and GPS. Our friends Alexa and Siri use AI to present us with the information we need. But what do we really mean by artificial intelligence? To understand AI, we first need to understand the concept of intelligence.
Many early AI efforts focused on the idea of ‘search trees’, which help to assess the effect of changes at any level and tracking the process that takes a process from the initial state to desired outcome. You can imagine search trees in action during a game of chess. With each move made, there are new options for achieving the outcome of a win. The notion of heuristics emerged in the 1990s to help inform search trees, and identify which processes are likely to have a positive outcome and which are more likely to result in a dead end.
In 1996, a computer called Deep Blue beat world champion chess player Garry Kasparov. Deep Blue could process 200 million possible moves per second, operating between six and eight moves ahead off the game.
In 1971 Terry Winograd created an interactive tool called SHRDLU. The tool responded to instructions in English, regarding the arrangement of boxes, blocks and other shapes using s simulated robot arm. SHRDLU could interact with rich dialogue that felt distinctly natural. However, this could be achieved only due to the constrained scenario to be resolved. Issues of intelligence, like perception, were not involved.
It is through our intelligence that we can handle the nuances and complexity of problems. As humans, we bring all of our years of accumulated knowledge into the assessment of a problem. We understand there are exceptions to rules- we understand the notion that birds fly, and at that same time that penguins are birds that are flightless. Operating a business and making the best choices requires the analysis of up to many thousands of different possibilities and combinations. Computers can try to use logic to solve problems but they do not have the same common sense reasoning that we do.
The modern era of AI has been focused on machine learning. Machine learning relates to learning from and making predictions based on data. Developing machine learning requires the processing and analysis of many sets of data.
The intent of machine learning is to build computers that do not need explicit instruction on how to handle every scenario they face.
Training and repetitive assessment must be carried out to equip the computer to handle this. In reinforcement learning, a computer is given the opportunity to assess whether the decisions it made through its thought tree processing had a positive or negative outcome. Decisions are repeatedly practiced to reduce the likelihood of a negative outcome. As sophisticated as machine learning may be at selecting the right decisions, it cannot explain that decision.
Deep learning is a term used to describe a system that can deal with complex problems. In 2014, Google acquired AI company Deep Mind, for a reported $400 million. They had created a system that could play arcade games - better than humans could, using reinforcement learning and machine learning to control the play. Later Deep Mind produced a system that could beat humans at the complex board game Go!
There is increasing pressure on businesses to use AI for data assessment and analytics. We hear so much about the rich data embedded in our work and are looking for the best ways to harness the power of the data. Most businesses have too much data and don’t know where to begin.
AI tools can help business identify opportunities and drive growth. They can help improve decision making and adapt to ever-changing scenarios and markets. There is also a risk of data bias or errors though. We are always going to need human reasoning to be able to see the full picture of what data is trying to tell us. Interpreting data requires both people who understand the functional and capacity of the data, but also people to understand the nature of the work being done to generate the data.
AI brings opportunities in just about every field of computing and applications. One emerging area is the use of AI in Enterprise Resource Planning (ERPs), a range of applications and systems that integrate business processes which had previously been handled separately.
The use of AI in ERP could radically change the way you handle business data and tasks. AI has the capacity to both augment and boost ERP systems, enhancing their capacity and improving accuracy and insights. AI can be used to enhance data analysis and expand the predictive capacity of ERP reporting. AI can do the heavy lifting in terms of data processing and number crunching.
There is enormous potential for AI to be able to analyze information generated by an ERP. AI can build on the integration activity of the ERP by offering opportunities for improvement to process and function. It does so by processing information in a much more expansive and efficient way.
Integrating AI into ERP also enhances predictive capabilities, enabling you to more readily anticipate future demands, risks, and opportunities. AI may also offer benefit by contributing to coding within the ERP, improving the speed and capacity of data integration through enhanced processes.
AI is bound to become a bigger component of ERP use.
You can gain a significant advantage by combining AI with your ERP. This process enables you to establish connections through data, which opens opportunities for data discovery. Using AI and machine learning in tandem with an ERP can add value to ERP activity.
AI processes and functionality are being refined and updated all the time. As with any technological change program, adding AI components to your ERP requires proper project planning and processes. Adding AI to an ERP is not as simple as flicking a switch or installing an integration tool. There should be a clear business need for adding AI to the ERP, and a clear set of measurable aims and objectives.
This means there may be additional considerations with resourcing and capacity. There are very few true AI in ERP technical experts around to facilitate the process. Adding AI to an ERP might also be just a band-aid solution if there is already a complex information infrastructure at work. It makes sense to investigate the long-term need for and approach to AI before attempting an incremental move to AI operations.
While AI has the potential to revolutionize how businesses operate, at this stage, AI could not replace an ERP system. AI offers additional value and add-on enhancements to enable you to make the most of your ERP. AI can improve data processing and adds intelligence to your ERP. Many organizations have their ERP tool at the heart of their operations; they are integral to day-to-day running and performance. It is unlikely that AI will replace the need for this work to be done.
Pimcore is a Master Data Management (MDM) tool that can deliver product management, digital asset management, and customer experience solutions. Pimcore can also act like an ERP in the way it spans multiple business functions and can bring different types of records together. Pimcore can be integrated with AI tools, such as OpenAI GPT-3, to enrich product descriptions. This can be useful if you have many hundreds of products and variations on your online shop or E-commerce solution.