Smart Solutions for Supply Chain Challenges: A Deep Dive into AI and CI

The impact of AI is undeniable and while it poses many challenges, supply chain optimisation is one area where the technology is having a positive impact. In the first in our series of insight articles Breathe’s Head of Software Delivery, Phil Rowlands explores its impact.

How does AI best fit in the supply chain?

The recent release and public engagement with generative AI and ChatGPT in particular has provided a pertinent reminder of how rapidly the capabilities of AI based solutions are improving. ChatGPT has proven to be quite disruptive in creative fields, or indeed any field that heavily relies on (previously human) written communication, but when it comes to the Supply Chain, many of its more difficult problems are not suitable for generative AI and a different approach is needed. Luckily AI is a diverse discipline with a wide range of tools, able to perform tasks that typically require human intelligence some of which are very much applicable to supply chain problems. Computational Intelligence (CI) is one of those tools – a subset of AI that specializes in problem-solving techniques inspired by nature and biology. It encompasses a range of methods, including evolutionary algorithms, fuzzy logic, neural networks, and swarm intelligence. These approaches are particularly well-suited for solving complex, dynamic, and uncertain problems—characteristics often encountered in supply chain management.

Supply Chain Challenges

Supply chain management is fraught with various challenges that make it an ideal candidate for CI-based solutions:

  • Uncertainty: Supply chains are affected by countless variables, including demand fluctuations, weather disruptions, and unforeseen events like the COVID-19 pandemic. CI techniques, such as fuzzy logic, can handle uncertain data and imprecise information effectively.
  • Optimization: Optimizing the supply chain involves making decisions about inventory levels, production schedules, and transportation routes. CI algorithms, such as genetic algorithms and simulated annealing, excel in solving complex optimization problems.
    Multi-objective Optimization: Supply chain management often involves multiple conflicting objectives, such as minimizing costs while maximizing customer service levels. CI algorithms can handle multi-objective optimization, ensuring a balanced approach to decision-making.
  • Adaptability: Supply chains must adapt to changing market conditions and unexpected disruptions. CI approaches, inspired by nature’s ability to adapt, can help supply chains become more resilient and responsive.
  • Complexity: Modern supply chains are highly intricate, with numerous interconnected components. Neural networks and swarm intelligence can model and manage this complexity effectively.

Applying Computational Intelligence

Computational Intelligence can be applied to each of these very real challenges faced by any organisation reliant on logistics and supply chain processes and while still in the early stages of its supply chain journey it has potential to make a real impact for the better

  • Demand Forecasting: CI techniques can analyse historical sales data and incorporate external factors to provide more accurate demand forecasts. Fuzzy logic, for instance, can handle imprecise data and generate probabilistic forecasts.
  • Inventory Management: CI algorithms can optimize inventory levels by considering factors like lead times, demand variability, and storage costs. Genetic algorithms can find near-optimal solutions to balance these trade-offs.
  • Route Optimization: CI can optimize delivery routes to minimize transportation costs and delivery times, taking into account real-time traffic data and route constraints.
  • Supplier Selection: When choosing suppliers, CI methods can evaluate multiple criteria simultaneously, including cost, quality, and reliability, to make informed decisions
  • Risk Management: CI can be used to assess and mitigate risks within the supply chain, helping businesses proactively address disruptions and vulnerabilities.
  • Pick Optimisation: Order sequencing in Batch Picking, and walking route optimisation are excellent candidates for nature inspired algorithms

While general AI has made significant strides in improving supply chain management, there are scenarios where Computational Intelligence shines even brighter. Its ability to handle uncertainty, multi-objective optimization, and complex, dynamic systems positions CI as a valuable tool for addressing the unique challenges of supply chain management. By embracing CI techniques, businesses can enhance their supply chain’s efficiency, resilience, and adaptability, ultimately gaining a competitive edge in today’s dynamic marketplace.

This evolution is just starting to have an impact on the real-life processes and challenges faced by many of those needing to improve and optimise their warehouse operation. Orquestr8 by Breathe Technologies is an AI based software platform that problems solves and optimises warehouse processes using the abilities of CI to simplify the integration of core warehouse functions and provide a harmonised eco system. It’s pick optimisation module is a great example of CI handling the interconnected processes and multiple objectives of the pick operation and making smart decisions to work out the optimal batching and route navigation that can reduce the pick walking distance by up to 50%.
Our recent picking article provides more insight into Pick Optimisation and key considerations when assessing opportunities to improve productivity, efficiency and accuracy.

As supply chains continue to evolve, the integration of Computational Intelligence is likely to continue to play a pivotal role in shaping their future success.