What are the main challenges and risks of implementing ERP AI for project management?

ERP projects inherently involve integrating at least 60% unique business practices in the implementation that require human intervention. AI is good for consistent and predictive transactions and data but normally, transactions involving signigicant financial impact need to undergo validations and discretionary handling. AI is totally against the "Different strokes for different folks" principle of quality management. There are critical human intuitive capabilities that cannot be replaced by AI.

Some ways I overcame this data quality problem was to design and implement a Data Quality team. It was composed of 10 employees that had to process thousands of asset inventory changes feeding SAP-PM. This team was implemented over 10 years ago and is still improving data quality today.

Project management relies heavily on accurate and reliable data. If the data integrated into the ERP AI system is flawed, outdated, or incomplete, it can result in inaccurate project insights and decisions. For instance, if project timelines are based on incorrect data, it can lead to missed deadlines and budget overruns.

A major issue for me is defining the scope and keeping the team focused on the business issues. What activities by the public are we trying to prevent (e.g. fraud) and what internal activities are we encouraging (e.g. ordering efficient quantities of goods / services)? Most importantly, does the data support the answers we seek or is there a risk that people will put too much confidence in the ERP reports? I never want to hear from a user, "we made that decision because the report told us to."

The main challenges of implementing ERP AI for project management include data privacy concerns, integration complexities, potential resistance from employees to AI adoption, and the need for skilled personnel to manage AI systems effectively.

Effective change management requires comprehensive planning, communication, and stakeholder engagement. Inadequate change management strategies, such as a lack of clear goals, insufficient resources, or limited stakeholder involvement, can lead to implementation delays, resistance, and ultimately, project failure. Investing in robust change management processes is essential for mitigating these risks.

Change management is key to a succesful ERP implementation project. Creating short videos and running an information campaign about the innovative ERP components being implemented will reduce resistance and maximize benefits.

Ao gerenciar a mudança de um sistema ERP, é crucial considerar tanto o impacto técnico quanto humano, garantindo que a infraestrutura de TI seja compatível e que os funcionários sejam devidamente treinados e apoiados durante a transição. Além disso, é importante estabelecer uma comunicação clara e contínua para gerenciar as expectativas e mitigar resistências, promovendo uma adaptação mais suave e eficaz.

Again, properly inform people about how ERP AI will handle sensitive information will create a feeling of safety for user communities.

I don't want just the *feeling* of safety. ERPs concentrate and link data. AI adds whole new layers of data analysis and pattern recognition. Everyone involved needs to be aware of the potential for misusing the data. Never before have ethics and internal privacy controls been so necessary. Give the Privacy Officer the support and tools they need to make sure the data is used only for authorized purposes.

Data Privacy and Security: ERP AI systems collect and analyze vast amounts of data, including sensitive project information and personal data of employees. Ensuring data privacy and security is essential to comply with regulations such as GDPR or CCPA. Mishandling or unauthorized access to this data can lead to legal penalties and damage to the organization's reputation. Bias and Fairness: AI algorithms may inadvertently incorporate biases present in the data used for training or in the design of the algorithms themselves. Biased decision-making in project management can lead to unfair treatment of employees. Organizations must mitigate bias through data preprocessing, algorithm transparency, and ongoing monitoring.

A inteligência artificial aplicada a sistemas ERP eleva eficiência e precisão operacional, mas requer rigorosa observância de práticas éticas, especialmente no que diz respeito à privacidade de dados e transparência nos processos decisórios automatizados.

Cost: ERP AI implementation can be expensive, particularly for small and medium-sized organizations. It can involve significant upfront costs, including hardware, software, and personnel. Additionally, ongoing maintenance costs and license fees may add up over time. Cybersecurity: AI systems can be vulnerable to cyber threats, such as hacking, malware, and data breaches. Therefore, organizations need to implement robust cybersecurity measures to protect sensitive data and systems. Overreliance on AI: Organizations may become overly reliant on AI-generated insights and predictions, leading to reduced critical thinking and decision-making skills. Therefore, it is essential to strike a balance between AI-generated insights & human expertise.

Make sure your business sponsor knows about tangible benefits about ERP AI. This will make project costs acceptable and will highlight the competitive advantages of implementing ERP AI.

AI requires in-depth knowledge of the processes impacted and type & quality of data being handled. I always compare it to a nanny who takes care of a child based on how she saw the mother did it (or from some thrown instructions) - the quality of caring will always be below par because she doesn't have that maternal instincts. With AI, how much will always prevail over how good - which will prove to be highly risky in the long run.