Leveraging AI in Operations Management: Opportunities and Threats

Artificial intelligence (AI) is a transformative technology that will fundamentally change the business landscape. In 2018, Google CEO Sundar Pichai went as far as to say that AI is more profound than “electricity or fire.” In a recent article published in the Financial Times, Pichai explained that “generative AI has captured the world’s imagination” and that “more and more start-ups and organisations are bringing AI-powered products and technologies to market faster than ever.” 

But as AI permeates various business sectors, a question arises: how does AI impact operations management? This article explores how AI will continue revolutionising operations management through process automation, data-driven decision making and predictive analytics. It will also explore the potential threats that come with AI, such as implementation challenges, data security concerns, the issue of AI hallucination and job displacement.

How Does AI Support Better Operations Management?

AI has the power to transform operations management for the better, primarily through process automation, data-driven decision-making and predictive analytics. We will take a closer look at how powerful AI systems can propel businesses forward.

Automating and re-engineering processes

One of the most transformative benefits of AI for operations teams is process automation. Repetitive, routine tasks such as data entry, inventory management and quality control can be partially or, in some cases, fully automated using AI, freeing up time for employees to work on higher-level tasks. 

Partial automation plays a role in the re-engineering of vital processes. For example, in an article exploring the impact of AI on processes, Dan Jeavons, who leads AI initiatives at Shell, wrote that the company is re-engineering work processes around AI and automation. Because AI brings new capabilities to business processes, companies need to rethink the tasks required, in what frequency, and determine who does them. 

To put this into context, at Shell, AI and automation are being used to reduce constraints on inspectors and maintenance technicians, with low-value-adding inspection tasks being carried out by drones and robots. Re-engineering processes around AI yields numerous benefits, including improved efficiency and accuracy, boosted employee productivity and significant cost reductions.

Data-driven decision-making

AI’s ability to sift through vast datasets and identify patterns will improve decision-making processes in operations management. Machine learning algorithms can uncover insights that might be missed or simply take too long for operations teams to find. These insights ultimately lead to better-informed decision-making, which enhances operational efficiency. For example, some companies use AI to analyse financial budgets and simulate financial scenarios, resulting in optimised spending. 

Another example of how AI is allowing businesses to make data-driven decisions is the optimisation of supply chain management. AI systems are able to identify optimal routes and predict potential disruptions to create a more resilient operation. For example, delivery company UPS uses an AI-powered tool to create the most efficient routes for its fleet. Since the initial deployment, it has saved UPS around 100 million miles and 10 million gallons of fuel annually. 

At Palladium, we allow our clients to make data-driven decisions through PrismGPT, a secure area we have developed, allowing companies to have better control over the information employees access and safeguard the confidential information fed into the platform. 

Predictive analytics

Predictive analytics, which uses data to forecast future outcomes, is a game-changer for operations management. The process utilises data analysis, machine learning, artificial intelligence and statistical models to find patterns in an attempt to predict future behaviour. Predictive analytics can be used to streamline operations, mitigate risk and boost revenue in almost all industries. Companies use predictive analytics models to manage resources and forecast inventory in order to operate more efficiently. 

For example, predictive analytics is used by manufacturing companies to predict when machines will require maintenance. Similarly, marketing teams use customer relationship management (CRM) tools to forecast customer behaviour and provide personalised product and service recommendations across platforms.

What Threats Does AI Pose to Operations?

It’s clear to see that AI systems are hugely beneficial to organisations across all sectors. However, businesses planning to optimise operations management with AI must understand and prepare for the threats involved. Implementation roadblocks, data security, AI hallucination and job displacement are some of the biggest challenges awaiting organisations planning to integrate AI into operations.

Implementation challenges

Successfully implementing AI in operations management is challenging and requires significant investment. Organisations must have the proper infrastructure in place, and employees need to be trained to work effectively with AI systems to optimise efficiency. There is much to consider, but following an AI implementation roadmap will guide business leaders through the complex process. 

Data security

AI systems rely on having access to vast sums of data—some of it sensitive in nature. AI systems utilised for operations management purposes require sensitive business information, including customer and supplier data. If this sensitive data is not adequately protected, it will be vulnerable to data breaches. 

Stories about AI-related data breaches have already hit the headlines. For instance, in March 2023, ChatGPT creator OpenAI revealed that the platform experienced temporary downtime due to a data breach. In a press release, the company published technical details about the problem, stating that a bug had caused some users to view other users’ chat history. 

But that’s not all. The bug also allowed some users to see other users’ first and last names, email addresses, payment addresses, credit card types, the last four digits of the card number and the expiration date. While OpenAI stressed that the number of users whose data was exposed was ‘extremely low’, the situation reveals that AI systems have weaknesses. To ensure data security, companies must ensure that robust security measures are implemented, and AI systems must be continuously monitored for potential weaknesses that could be targeted by bad actors.

Hallucination

One threat that AI poses is hallucination. Rich Klee, Director of Product and Tech at Palladium, explained that AI systems are “prone to hallucinate with careless prompting and can provide inconsistent results.” AI hallucination occurs when AI systems, particularly those that utilise deep learning, generate or interpret data based on their training incorrectly. The AI system may overgeneralise, identify patterns that do not exist, or wrongly interpret ambiguous data.

In operations management, AI hallucinations could have serious consequences. For example, in the context of equipment maintenance, hallucinations could lead to costly, unnecessary maintenance or missed opportunities to prevent equipment failure. In logistics, for instance, AI systems might hallucinate patterns in delivery times or routes that do not exist, which will significantly reduce operational efficiency. As a result, AI systems should be continuously monitored and updated to reduce the chance of hallucination. For example, regularly comparing the output of AI systems with real-world data will allow users to identify hallucinations.

Job displacement

One of the most commonly discussed threats is job displacement. Daren Acemoglu, a professor and labour economist at the Massachusetts Institute of Technology, says automation, including generative AI, could “increase productivity while reducing wages and employment.” While job displacement is a worry, employees can leave some of the more repetitive, tiresome tasks that humans do not generally enjoy doing, such as data entry, inventory management, quality control and scheduling coordination, to AI. To mitigate job losses, businesses should reskill and upskill employees, allowing them to remain in the business but in a different capacity, leaving routine tasks to AI.

In Summary: A Promising Future

There is no doubt that AI has a promising future for businesses operating in all sectors. The technology offers significant opportunities for process automation, data-driven decision-making and predictive analytics, improving efficiency and ultimately boosting bottom lines. However, AI also poses several challenges that businesses must address. Data security, AI hallucination, job displacement and implementation challenges are of particular concern, and failure to address these threats could have disastrous consequences. Therefore, to ensure AI is implemented responsibly, organisations should follow a carefully formulated strategic roadmap that maximises the benefits while mitigating the risks.