Not Everything can be Automated. Here’s Why.

While AI systems are now capable of revolutionary things within pre-determined parameters, they do not possess the adaptability, personality, or empathy required to handle freight anomalies.

Sutthiphong Adobe Stock 566449343
Sutthiphong AdobeStock_566449343

It has been one year since ChatGPT introduced itself to the world. Since then, one thing has become clear: artificial intelligence (AI) will soon be involved in many aspects of our lives. This is already true in the business world: AI has become a focal point for businesses across every sector, with uses ranging from customer service chatbots to algorithms that identify and suggest efficiency improvements.

As this becomes an evident reality, many worry about the future. Will AI take over human jobs? Will certain positions become obsolete? Jobs with highly repetitive actions and a clear process logic could very much be at risk, but these positions were already under pressure from existing machine learning, chatbots, and robotic process automation. This new breed of generative AI can perform impressive analytical tasks without human oversight, but when it comes to making real-world business decisions, concerns regarding data accuracy and any number of outside factors, making the likelihood of a full replacement unlikely. The more realistic outcome is a partnership between human and AI systems, allowing people to take advantage of AI’s capabilities in data processing and calculations while maintaining a human being’s adaptability and social skills.

In the world of logistics and supply chain, AI is currently being implemented in several ways. Route optimization is one example where AI can far outperform a human mind. It can gather, analyze, and evaluate massive datasets extremely quickly. This makes AI route optimization attractive for logistics companies. Consolidation is another service that AI can assist with, quickly identifying opportunities to use fewer trucks to move the same amount of freight by organizing loads around multiple stops. AI is also utilized in warehouse stocking, inventory availability, and cost predictions.

Notice a pattern: in the supply chain space, AI has mostly been used as a tool for data processing and predictive analytics. While these are useful applications, there is still massive potential for more AI involvement in the life of a load. Many experts believe that the next phase of AI involvement in the supply chain will be in appointment scheduling.

Appointment scheduling is also a headache for operators. No two delivery or pickup locations use the same scheduling system. Some facilities use portals to schedule appointments while others schedule through email. Many use a combination of both. Some are strictly organized while others are not. Appointments must be made correctly, giving drivers enough time to arrive while still beating the load’s due date. If an issue arises during transit, appointments must be rescheduled around the availability of dock space and labor to unload the truck. Scheduling appointments for loads is dull, time-consuming work that requires attention to detail and clear communication between people.  Robotic process automation has assumed some of the grunt work here, in many cases removing the human effort from logging in and scheduling specific appointments in many disparate systems, but this still requires decision-making human intervention, as well as a massive IT effort to create and maintain connections to an ever-growing number of systems.

If AI could take over scheduling appointments, it would free up a lot of time that human operators currently spend on dull tasks. So, why aren’t bots scheduling appointments already? The answer lies in the many different systems described above. Every location requires different information and processes to schedule, see available appointments, and confirm after requesting an appointment. This makes it exceedingly difficult to train an AI system to take over scheduling, especially for a broker or carrier that deals with many different facilities.

The Scheduling Standards Consortium, a new coalition between supply chain companies, aims to change that. The organizations are working together to set a standard for scheduling appointments. If they are successfully able to pull the fragmented processes, information requirements, and scheduling portals together, it would be a massive step toward AI systems taking on scheduling responsibilities.

Still, the road to scheduling standards is a long and winding one - smaller carriers, shippers and receivers may not have the technical skills or available budget to implement or upgrade systems. The largest organizations may be resistant to cede control over processes which may currently be a competitive proprietary advantage. Finally, the difficult freight market has seen large digital brokerages, who were large drivers of standards creation, go belly up.

As bot scheduling becomes possible, AI will take a step up from its current position as a tool towards its new position as a partner alongside operators. When organizations can train their AI to respond to external factors, adjust appointments as issues arise, and notify drivers quickly, they will begin to harness the true potential of AI: utilizing its natural language and data analysis capabilities simultaneously.

Despite the potential, however, it is not realistic to think that AI will take over most human positions soon.  The supply chain, whether examined on a global scale or the local business level, is simply too unpredictable to rely entirely on AI to handle any external factors that come up. Any supply chain professional is acutely aware that problem-solving is often not as simple as, “Truck won’t make its delivery appointment, reschedule the appointment, notify driver, problem solved.” A problem so simple is quite rare. Often, problem-solving requires knowledge of previous solutions, human relationships, and nuanced conversations with warehouse operators, drivers, and dispatchers.

Issues like angry or unreachable drivers and car accidents are issues that require human intervention to solve. When dealing with these problems, AI will return to its “tool” role: analyzing alternate routes, suggesting replacement carriers, or providing other pertinent information to its human counterparts who are having conversations to sort out the situation.

In the freight world, the problems you can imagine are only a fraction of issues that can and will arise. While AI systems are now capable of revolutionary things within pre-determined parameters, they do not possess the adaptability, personality, or empathy required to handle freight anomalies. Peak efficiency will be achieved through a team of AI and human supply chain experts, not through replacement. Zero human involvement is simply not possible for the foreseeable future, due to this simple fact: not everything can be automated.