AI wants to turn ‘ugly freight’ beautiful, but bad data could mar its efforts 

Just as the internet and the cloud did before, AI has reached an inflection point. While the technology’s usage has doubled since 2020, business leaders’ concerns around its cost have increased by more than 14 times this year alone. With the present AI reality marked by cost overruns and inaccuracies, corporate decision-makers are justified in their feelings of unease, particularly in the economical and meticulous logistics space.  

As AI’s hype cycle evolves and matures, it’s increasingly clear that the path to successful implementation intersects three key factors: 1) AI adoption by carriers of oversized, larger-than-parcel items — or “ugly freight” in industry parlance, 2) high data quality for training and inferencing applications, and 3) customer contributions — at scale.  

Boosting AI adoption 

Last year, AI advancements in logistics seemed destined to transform the parcel space, led by many of the same industry players who normalized one-day delivery during the e-commerce boom. Amazon asserted that its AI tech would “make holiday shopping seamless,” UPS claimed the technology could combat porch pirates, and FedEx promised more accurate delivery estimates. Yet, the technology’s actual impact remained in question. Only one in four parcel shippers are actually using the technology today, and the industry as a whole remains sluggish in its AI adoption.  

Enter “ugly freight,” the most compelling use case for AI in logistics, and one that thus far has seen little exploration. Although ugly freight like an antique furniture piece or a vintage motorcycle can be beautiful, moving it rarely is. It can be frustrating, expensive, and prone to damage. For these reasons, any AI advancements in this space are far more valuable than in the already-well-oiled parcel shipping sector. Non-conveyable oversized items are nearly impossible to automate, because moving them brings distinctive challenges. The weight, irregular size, and fragility of many ugly freight shipments all introduce unique variables into the shipping process. These unique attributes can also slow shipping times and introduce additional costs for services such as specialized white-glove delivery to prevent item damage and ensure a positive customer experience.   

The situation isn’t that much better for shipping customers of ugly freight, who can face numerous issues if an item’s dimensions or weight are not calculated correctly. A price estimate can prove wildly inaccurate, or an unprepared carrier could be unable to fulfill a shipment. In either case, discrepancies between different items of similar shapes and sizes (a couch built in 2020 is vastly easier to move than a sleeper sofa from the 1980s) can escape a carrier’s eye with serious consequences. AI could do better if it’s equipped with high-quality data.  

Ensuring data quality 

Recent advancements in deep learning algorithms have vastly transformed AI’s ability to conduct complex visual recognition tasks, with Google’s Lens technology now processing over 20 billion searches a month with a 92.6 percent accuracy rate. If the same technology could be applied to larger-than-parcel shipping, it could fill gaps that carriers may otherwise miss and may even amend a user’s inaccurate dimensions or weight calculations for a better, more accurate quote. Yet, this accuracy is contingent on strong underlying training and inferencing data.  

Unfortunately for the logistics space, this data seems easier to find than it ultimately is. Despite mountains of proprietary historical pricing and shipping data, most of this data isn’t properly structured or suitable for AI training. A recent study found more than 45 percent of newly created data records fall into this category, with at least one critical fault. 

 Before utilizing datasets for model training (or evaluating the dimensions of an ugly freight piece), logistics pros must first comb through their training data for missing values, outliers, or other abnormalities that may cause an AI model to return inaccurate results once online. These values can be replaced with high-quality synthetic data, but only if they’re caught before model training. This is why data validation is the most time-consuming step in AI development, and why shoring up one’s data warehouse is the most critical step to leverage AI. Ugly freight shipping may still be transformed by the power of AI, but only if quality data is at the core of every algorithm used.   

Scaling customer contributions 

Even if a model is trained on the highest-quality dataset, its knowledge is still finite. Since every larger-than-parcel item is unique, AI may successfully compile the general attributes of an item, but any dimensions generated will be ballpark figures. That’s why customer input is the final piece of the puzzle to creating an accurate algorithm for ugly freight shipping.   

With every armoire and ottoman shipment input into an AI tool, the data used to estimate future shipment dimensions and weight will improve. Add in context inputted by the customer for each shipment (differentiating a sleeper sofa from a couch, or a 20th-century piece from a fast-furniture special), and an algorithm is well on its way to becoming incredibly accurate over time.   

AI’s expansion into the logistics space has been constrained thus far by a myriad of factors, chief among them being high cost, perceived complexity, and the technology’s current inability to deliver on years of lofty promises. Despite the industry’s trepidation and muddled forecasts of the technology’s future impacts, however, the AI revolution is still coming to the logistics space. If ugly freight shippers embrace AI, algorithms are trained on high-quality data, and customer input is ongoing, AI can and will transform transport.  

For a list of the sources used in this article, please contact the editor.  

Jami Caruso
www.uship.com 

Jami Caruso is the VP of Customer Operations and Home Delivery at uShip, the world’s largest and most trusted online transportation marketplace. People and businesses use uShip because it’s easy and affordable to price, book, and ship everything from cars to cranes, freight to furniture, and households to horses—whether it’s going local or long distance.