UPS & Automation
We live in a world that increasingly relies on innovations in technology to produce the most efficient, rapidly gratifying outcomes. The interdependence of technology and our daily lives has never been more evident than within the package delivery industry.
The package delivery industry has seen an unsurmounted increase in demands since the nationwide shutdown that led to the closures of many businesses. As a UPS seasonal helper this past summer, I witnessed the cutting-edge process of the ORION system, which UPS uses to expedite their deliveries.
Developed by UPS, ORION uses advanced algorithms, AI, and machine learning to determine the most efficient package delivery route, saving the company both time and fuel costs.
Drivers have a portable device that displays the route and signals the driver when and where to stop at various destinations. Since its development, the ORION system has saved UPS over 100 million miles of extraneous driving and 10 million gallons of fuel. On paper it has made UPS’s delivery process far more efficient, but can machine learning truly replace human experience?
I found my experience with ORION to be quite different. As a driver helper, I worked with multiple drivers on different routes in a mid-sized urban area. I sat alongside the driver and delivered the packages door to door.
Through this, I was able to compare the effects of the automated GPS system on the drivers’ efficiency. During my first shift, I was assigned to a heavily settled urban area. The driver I worked with that day was brought upon from a different sector in the company to help with the COVID delivery rush. He followed the ORION system religiously, but throughout our route, he made repetitive turns, often passing places we would have to circle back and deliver to later.
A few weeks later, I was relocated to a delivery route in a more rural area. This route saw upwards of 200 stops a day, and the delivery locations were far more spread out than my previous route.
The driver I worked with on this day was far more seasoned than the last; he had been working as a UPS driver for over 20 years and had been assigned to this particular route for the last 5. He knew every street and detour in the town and was friendly with all of its residents. He also did not follow the ORION system whatsoever. He instead went off of instinct, knowing which neighborhoods would be busiest and when to avoid them, how likely someone would be to be home for a signature, and countless ways to work out the kinks of the route that often slow drivers down.
There was one week when this particular driver was on vacation, but I remained on the route with his substitute. This driver was familiar with this route, but uncertain how it had changed over the years, so he followed the ORION route. Despite the COVID rush slowing down and seeing less stops per day, the route that normally took 6 or 7 hours to complete with the initial driver took 8 hours that day.
This first hand experience with the effects of UPS’s efficiency algorithm on various drivers led me to the question: to what extent can technology replace the invaluable instinct of human experience?
Less experienced drivers who knew no better relied on ORION to get them through their routes, but those who solely followed their experience and instinct completed their routes much quicker. I found through my time with UPS that drivers who use ORION operate much slower than those who do not. This led to my further reading on the integration of the ORION platform into the drivers’ lives. The drivers who used ORION took a longer time to complete the route compared to those who were familiar with the area, and used their own instinct, judgment, and knowledge to complete the route.
UPS’s ORION is known to have successfully optimized delivery routes, thus saving company money.However, drivers who do not utilize ORIONview optimal efficiency as the least amount of time it takes them to complete their routes. With the reliance on ORION to streamline the daily services we rely on, there remains a disconnect between UPS’s technological advancements and the needs of their workers on the ground highlights a new problem: how do we get people and machines to work together?
Using ORION saves UPS money, but relying on instinct gets drivers home quicker. Is there a way to do both?
AI, algorithms, and machine learning provides the optimization of aspects of our daily lives, but at UPS and similar companies, it undeniably removes the autonomy of the worker. ORION ran into problems in its initial development with disparities between the drivers’ experiences and the outcomes of the algorithm, requiring significant modifications in 2007. While improved, these disparities are still prevalent 13 years later, highlighting what can happen when companies take a top-down approach to streamlining their work.
There is always work to be done with the ever-developing world of AI, and it would benefit companies such as UPS to explore a bottom-up approach in modifying their algorithm to save time and money, while also improving the drivers’ experiences.
Written by Catherine Hadro
Edited by Christine Lee & Alexander Fleiss