Navigating Digital Transformation: When to Use AI and When to Stick to Simpler Solutions

By Thomas Kiessling, CTO Siemens Smart Infrastructure & Gerhard Kress, SVP Xcelerator Portfolio Siemens.

  • 2 weeks ago Posted in

All businesses are to some extent walking the sustainability tightrope and almost all are grappling with cost management and an increasingly complex regulatory environment. These challenges are putting renewed emphasis on the need for digital transformation – especially of outdated manual processes. However, as we navigate this transformation a critical question arises: when is the deployment of advanced technologies like Artificial Intelligence (AI) truly necessary, and when should we rely on simpler more established solutions? 

The promise of AI is immense. It purports to revolutionize everything from supply chain management to customer service. But it’s essential to remember that AI is not a silver bullet. While it excels in handling complex tasks and vast datasets, AI also comes with significant costs—both financial and environmental—that can outweigh its benefits if not carefully considered. 

The Complexity of Digital Transformation 

Digital transformation involves more than just adopting new technologies; it requires a fundamental shift in how businesses operate and deliver value to their customers. This transformation is particularly challenging for industries with established infrastructures—often referred to as brownfield environments—where older systems and hardware need to be integrated with new digital tools. 

Consider the transportation industry as a prime example. The sector faces the dual challenge of reducing carbon emissions while meeting growing demand. Digital tools, such as Internet of Things (IoT) devices, can provide real-time data on vehicle performance, traffic conditions, and energy consumption. This data can be used to optimize fleet management, reducing costs and environmental impact. However, the question remains: should this optimization be driven by AI, or can simpler, rule-based systems achieve the same results at a lower cost? 

Rule-Based Algorithms vs. AI: A Strategic Choice 

Rule-based algorithms operate on a set of predefined instructions and are particularly effective for straightforward tasks with limited variability. They are cost-effective, require minimal computational resources, and are easy to implement in environments where the complexity is manageable. These systems have been the backbone of industrial automation for decades, providing reliable performance with minimal risk. 

On the other hand, AI-based algorithms are dynamic, capable of learning from data, and can adapt to new information or changes in the environment. This makes AI invaluable in scenarios that involve pattern recognition, prediction, and decision-making in unpredictable environments. However, AI’s ability to handle complex tasks comes at a price—both in terms of initial investment and ongoing operational costs. 

For example, when managing an electric vehicle (EV) fleet, a rule-based system might schedule charging during off-peak hours to minimize costs. This approach is straightforward, predictable, and effective in many cases. However, if the goal is to optimize charging based on real-time data, such as fluctuating energy prices, weather conditions, and grid capacity, AI can offer a more sophisticated solution. Yet, the question remains: is the additional complexity and cost of AI justified in this scenario? 

The Environmental and Economic Costs of AI 

The deployment of AI can come with significant environmental and economic costs. Indeed, the costs can rack up very quickly. AI algorithms, particularly those involved in deep learning, require substantial computational power, leading to higher energy consumption. This is particularly concerning given that data centres—where AI computations are often carried out—are already significant consumers of electricity. As businesses push towards ambitious sustainability targets, the energy demands of AI has the potential to undermine these efforts. 

Moreover, AI’s reliance on large datasets presents its own challenges. Not all industries have access to the volume and quality of data required to train AI models effectively. In such cases, businesses may find themselves investing heavily in AI infrastructure without realizing the expected returns. 

Microsoft and Google, for example, have both reported increases in their carbon emissions due to the energy consumption of their AI-driven services. This raises a critical question for businesses: how do we balance the promise of AI with the need for environmental responsibility? 

When Simpler Solutions Are More Effective 

Given the costs and complexities associated with AI, there are many instances where simpler, rule-based solutions are not only sufficient but preferable. These include scenarios where: 

- Task complexity is low: For routine, well-defined tasks, rule-based systems can deliver reliable results without the need for AI’s advanced capabilities. 

- Data is scarce or inconsistent: AI thrives on data, but in environments where data is limited or of poor quality, rule-based systems can provide more consistent performance. 

- Transparency is required: Rule-based algorithms are inherently transparent, making it easier for businesses to understand and explain their operations. This is particularly important in regulated industries where decision-making processes need to be auditable. 

- Cost is a critical factor: The lower implementation and operational costs of rule-based systems make them an attractive option for businesses looking to control expenses. 

For instance, a rule-based approach might involve charging electric buses during off-peak hours. This is a straightforward solution that doesn’t require the complexity of AI. However, as the system scales and as more variables are introduced—such as the need to integrate renewable energy sources or to balance grid loads in real-time—the benefits of AI become more apparent. 

When AI is Indispensable 

Despite the potential drawbacks, there are situations where AI’s capabilities are indispensable. These typically involve: 

- High complexity and low error tolerance: In scenarios where precision is critical such as managing emergency services fleets (e-ambulances) — AI’s ability to process vast amounts of data and make real-time decisions can be invaluable. 

- Dynamic environments: In industries where conditions change rapidly, such as logistics or finance, AI’s adaptability allows businesses to respond to new challenges more effectively than rule-based systems. 

- Unstructured data: AI excels in processing unstructured data, such as images, videos, or text, which rule-based systems cannot handle. This capability is increasingly important as businesses look to leverage all available data to drive insights. 

For example, AI can optimize charging strategies for EV fleets by analyzing real-time traffic patterns, energy prices, and grid conditions. This dynamic approach allows businesses to not only reduce costs but also support grid stability and contribute to broader sustainability goals. 

The Right Tool for the Right Job 

Digital transformation is not a one-size-fits-all journey, and neither is the deployment of AI. It is essential to carefully consider when AI is truly necessary and when simpler solutions can deliver the required outcomes more efficiently. 

By strategically deploying AI where it adds the most value and relying on rule-based systems where they are sufficient, businesses can navigate the complexities of digital transformation while controlling costs and minimizing their environmental impact. The key is not to get caught up in the hype of AI but to focus on using the right tool for the right problem.   

By Krishna Sai, Senior VP of Technology and Engineering.
By Danny Lopez, CEO of Glasswall.
By Oz Olivo, VP, Product Management at Inrupt.
By Jason Beckett, Head of Technical Sales, Hitachi Vantara.
By Dael Williamson, Chief Technology Officer EMEA at Databricks.
By Ramzi Charif, VP Technical Operations, EMEA, VIRTUS Data Centres.
Companies are facing a Catch 22 when it comes to the need to invest in new forms of AI, whilst...