In 2021 Statista.com calculated the volume of data created, captured and consumed worldwide had hit 79 zettabytes (that’s 79 with 15 zeros) and was projected to more than double by 2025. With 2025 just around the corner, it is without question that data is a precious asset to every business but turning it into a strategic competitive advantage remains out of reach for most. Doing so requires keeping up. Keeping up with the sheer volume, frequency, evolving formats and required skills to source, ingest, transform, and process data. This simply cannot be economically solved using traditional approaches.
The solution is obvious (and no longer a revelation), we must deploy a mix of Artificial Intelligence (AI), Machine Learning (ML) and Automations into our existing business operations. What is less obvious is where and how organisations can leverage these technologies to improve decision-making and provide insight to achieve strategic advantage. The three examples outlined below are demonstrated applications of AI/ML to solve some of the most common and complex problems in the Transport Operations domain (i.e. the operations of shippers and transport service intermediaries).
Use Case 1: Identification of Invoice Anomalies
The transportation industry is rife with invoicing errors, with an average error rate of 8 to 10%. These mistakes affect cash flow and profitability, but they also have the potential to damage reputations if they are shared with clients. Conventional error mitigation techniques are not scalable since they depend on labor-intensive procedures and extensive domain knowledge. With the use of sophisticated ensemble models and/or statistical outlier detection, machine learning can provide a scalable approach to anomaly detection. Businesses can manage by exception, reviewing and actioning only those consignments flagged as anomalous every billing cycle or even automate the end-to-end credit request process with the use of automated flows and agents.
Use Case 2: Dynamic Pricing
Uber’s real-time fare adjustments serve as an example of dynamic pricing, which is crucial for balancing supply and demand in the transportation sector. Unlike Uber, traditional transport contracts are typically fixed for 1-3 years, making deviations from expected demand costly. Multi-year contracts provide cost certainty but fail to account for dynamic market changes. ML models augmented with external indices and lead indicators can enable proactive identification of changes in demand profiles, facilitating more frequent price recalibrations. This approach can minimise margin erosion and align sale prices with costs and expected margins. The ability to pass on increasing costs as soon as possible can change the profitability (and in turn viability) of some customers.
Use Case 3: Revenue/Margin Leakage
Margin erosion poses a significant challenge in the freight industry, often requiring extensive efforts to identify and rectify root causes. ML can address this challenge by analysing historical pricing data to identify deviations from expected margins at the transaction level. Regression-based algorithms can be deployed to isolate pricing components which impact margins, pinpointing the source and magnitude of margin erosion.
Conclusion
Implementing ML solutions in shippers’ transport operations requires careful consideration of the specific data context, normalization, feature selection, and external data augmentation. A comprehensive understanding of domain-specific challenges and ML techniques is essential for delivering accurate automatable solutions that can provide tangible benefits.
Whilst ML adoption requires some effort, the benefits of reduced cycle times, improved decision-making, and operational efficiencies can free up staff to work on higher value tasks and can be invaluable in establishing and maintaining data as a competitive advantage.
IntrinsiQ Logistics 
Leave a comment