How AI in Logistics Could Affect Component Lead Times

The application of Artificial Intelligence (AI) to the complex process of scheduling truck driver load assignments promises to boost the efficiency of shipping logistics, as well as demand for AI chips
like graphics processing units (GPUs).

The trucking industry is now contending with a paucity of workers, with an 80,000-driver shortfall
that is expected to double by 2030, according to the American Trucking Association (ATA). However, another solution exists to address this challenge beyond simply hiring more drivers.

How AI Can Optimize Existing Logistics Workers’ Time

Increasing the efficiency of today’s workers, could present a solution to this growing problem. More than 15% of all miles driven by truckers in the U.S. have no freight, according to the American Transportation
Research Institute. As a result, the trucking industry spends $288 billion annually to move empty cargo space, according to an estimate devised by scheduling software company Optym.

This challenge would appear to be readily solvable by using computer technology to improve scheduling efficiency. However, attaining optimal driver scheduling requires dealing with a mathematical issue that has challenged experts for years: the traveling salesman problem. The traveling salesman problem involves finding the shortest route possible when visiting multiple cities. Like other optimization problems, the traveling salesman problem is challenging to solve, and calculation grows vastly more complex as more destinations are added.

With all the variables involved in the problem, traditional computing approaches calculate one new route at a time, requiring the processing of a huge number of variables.The neural networks at the heart of today’s AI systems are more suited to solving problems like this.

Networks can be trained to take on different structures, allowing them to tackle problems like the traveling salesman problem with greater accuracy and efficiency than traditional computing approaches. As a result, AI is likely to play a significant role in improving trucking efficiency in the coming years.

Translating This Trend into Future Demand

Applying AI to yet another challenge will continue to drive demand for AI processing horsepower in
enterprise data centers and cloud computing operations. This, in turn, will propel sales of AI processors designed to accelerate neural networks.

The global market for data center AI processors is set to soar to more than $79.6 billion in 2028, nearly five times the $13 billion total in 2022, according to a forecast from Supplyframe Commodity IQ. However, since the formulation of this forecast, new events like the rise of ChatGPT have caused market expectations to inflate. 

This factor is contributing to tighter supplies of AI processors. Lead times are contracting for the ASIC/chipset category, including GPUs and other AI accelerator chips. Nearly 70% of at-volume lead times for ASIC and chipset parts were at greater than 36 weeks in the second quarter of 2023, down from 87% in the fourth quarter of 2022, according to the Commodity IQ Lead Time Index.

But as applications like logistics management translate into higher demand, lead times and overall availability will suffer. For electronics industry supply-chain participants, the application of AI
to truck scheduling will be a double-edged sword, improving logistics efficiency, but exacerbating challenges in sourcing key AI processing chips.