EmailMeForm



Delivering the Future: Examining the Impact of Autonomous Vehicles on Supply Chain ManagementA Capstone Project Submitted in Partial Fulfillment of theRequirements of the Renée Crown University Honors Program atSyracuse UniversityBenjamin J. HouleCandidate for Bachelor of Science Degreeand Renée Crown University HonorsFall 2019Honors Thesis Project in Supply Chain ManagementCapstone Project Advisor: _______________________Zhengping Wu, Associate Professor of Supply Chain ManagementCapstone Project Reader: _______________________ Gary La Point, Professor of Supply Chain Practice Honors Director: _______________________ Danielle Taana Smith, Director Abstract Supply chain management consists of the activities required to plan, control and execute the flow of a product as efficiently and effectively as possible. Managing transportation and logistics are critical to a supply chain’s success. Collectively, firms spend over $3 trillion a year on transportation and it represents a key focus area for firms to reduce costs. The introduction of autonomous vehicles will have a significant impact on how supply chain managers work and the strategies they use. Autonomous trucks are expected to help reduce lead times and decrease transportation costs. This will ultimately reduce a firm’s inventory levels and minimize the influence of the bullwhip effect. Given decreased costs and faster resupply, it will be more economical for firms to make smaller orders, more often. Automatic guided vehicles in warehouse settings reduce labor costs, improve worker safety, and decrease damaged inventory. Despite these benefits, there are several areas of concern regarding the implementation of autonomous trucks and vehicles. Infrastructure to support them needs to be developed and could prove to be expensive. Autonomous guided vehicles supplement labor in warehouses, and pay themselves off in less than a year, on average. They also increase worker safety and reduce damaged goods by limiting process variation and eliminating elements of human error. Autonomous trucks and autonomous guided vehicles would also alter the way many people in the transportation and warehousing sectors work, which may cause some delay in adoption by those hesitant to change. Because autonomous trucks are so early in their lifecycle, data are relatively limited. Further research will need to be done in the future when the results of their implementation are documented.Executive Summary: Supply chain management (SCM) is the broad range of activities required to plan, control and execute a product's flow, from acquiring raw materials and production through distribution to the final customer, in the most streamlined and cost-effective way possible (Rouse, 2018). Transportation and logistics are a critical part of SCM. Collectively, firms spend over $3 trillion a year on them, making them a prime target for cost minimization. Autonomous trucks and automated guided vehicles present an opportunity for managers to cut their costs. Autonomous trucks (ATs) and automated guided vehicles (AGVs) use technologies such as GPS, radar, sonar, lasers and cameras to gather information on the vehicle’s surroundings. That data is then used as inputs for an artificially intelligent “brain,” which uses the information to make decisions on speed and direction, while staying within the rules of the road. Artificial Intelligence (AI) is most successful when performing long, repetitive tasks with a strict ruleset, making long-haul driving an opportune application. The benefits that come alongside transitioning to ATs and AGVs are impressive. Replacing a human driver with an autonomous truck may decrease the time it takes to travel from New York City to Los Angeles by up to more than 50%. Decreasing lead-times make it more beneficial for supply chain managers to place smaller orders more frequently, improving the reactivity of the entire chain and freeing up capital that was previously tied up in inventories. The reduction will also lead help mitigate the influence of the bullwhip effect, a phenomenon that has decreased supply chain efficiency, which would result in even greater cost savings. AGVs decrease labor costs in the long run and significantly improve worker safety. Firms that invest in AGVs will begin to see positive returns on labor costs between 45.8 and 68.8 weeks, on average. AGVs also eliminate human performance variations, leading to increased personnel safety and a reduction in damaged inventory. (Citation?)Citation? However, these benefits do come with significant costs. Ports and fulfillment centers will need to switch to 24/7 schedules, the roles of drivers will transition to more “support” jobs and infrastructural improvements must be made to support the autonomous trucks. Warehouses will also need to develop infrastructure to support AGVs. Additionally, there are ethical concerns surrounding the replacement of some workers with AI and the shift of workforce demands that will follow. These, among other factors may slow the adoption of autonomous trucks. Table of ContentsAbstract2Executive Summary3Acknowledgements6Introduction7What are Autonomous Trucks, and How Do They Work?9A Brief History of Artificial Intelligence9Autonomous Trucks and How They Work10Applications to Supply Chain Management12Impact on Multi-Period Inventory Models13 Mitigating the Bullwhip Effect17 Industry Disruption18Case Study19Automated Guided Vehicles22Introduction to Automated Guided Vehicles and Warehousing22How Automated Guided Vehicles are Effective25Issues and Potential Threats29Conclusion32Citations34AcknowledgementsI would like to thank Professor Zhengping Wu and Professor Gary La Point for agreeing to be my advisor and reader. Your knowledge and advice both in and out of the classroom was crucial to my success completing this project. I would also like to thank Professor Steve Sawyer, whose class inspired me to explore this topic, and provided the groundwork for what would eventually become my thesis. Finally, massive thanks go out to my friends and family for supporting me throughout this process.IntroductionTechnology is advancing quicker than ever before. Technological developments, such as autonomous trucks (ATs) may be disruptive to many industries and alter how they function. One of the keys to industrial success is a responsive supply chain; products need to get from place to place as efficiently as possible. Autonomous trucks are going to have a significant impact on inventory management and decrease lead times and costs. However, there are several ethical and economic issues that arise. The roles truck drivers play will significantly change. Shifting demands in the workforce will orient truck drivers towards technological support functions and last-mile deliveries. Infrastructure to support ATs will need to be developed, which may be costly. What type of infrastructure will be required? Examining the impact of autonomous trucks, measuring their costs and benefits and analyzing how they will change the way people work is critical for successful adoption and integration. Supply chain management (SCM) is the broad range of activities required to plan, control and execute a product's flow, from acquiring raw materials and production through distribution to the final customer, in the most streamlined and cost-effective way possible (Rouse, 2018). Logistics and transportation are a critical part of supply chain management; goods need to get from point A to point B as quickly and efficiently as possible. The supply chain management program at Syracuse University, the first program of its kind in the United States, was initially dedicated to solely study transportation, eventually adopting the other aspects that characterize the field. Is this comment needed or relevant to your topic? It seems to be out of place with everything else. Firms spend $1.0093 trillion every year on transportation costs, including the costs of shipping goods and the costs of vehicles and maintenance. Citation? Additionally, the value of time spent by firms on transportation totals $2.1723 trillion, combining for a total value of over $3 trillion a year (Winston, 2010). This presents a prime area of opportunity for firms to cut down on time and save costs. The utilization of artificial intelligence in autonomous vehicles, such as trucks, may be the solution managers have long sought after. Autonomous trucks will have enormous impact on the way supply chains are managed and will certainly alter the strategies managers will employ. First (and most obviously), the gradual elimination of human drivers will lower labor costs significantly and bypassing the driver’s required breaks will reduce lead times. Additional technology keeps trucks a certain distance apart, maximizing fuel efficiency and further reducing costs. These benefits are just the surface of many that could be realized with this introduction, however, significant changes to the industry are on the horizon. Autonomous trucks will totally alter the function of “driver”, change the way managers operate and strategize and would require significant upgrades to current infrastructure to support these developments. Despite these changes, McKinsey and Co. estimates that fully autonomous vehicles will save the trucking industry between $85 billion and $125 billion. Another study by Clements and Kockelman estimates that if autonomous vehicles capture a large share of the market, the overall economic impact would be about $1.2 trillion, or $3,800 per American, per year (Clements and Kockelman, 2017). The purpose of the paper is to holistically examine the impact autonomous trucks and automated guided vehicles will have on Supply Chain Management and the work that managers do, as well as provide analysis on the costs and benefits of these changes.What are Autonomous Trucks and How Do They Work?To fully comprehend the effect autonomous trucks will have on supply chains, it’s critical to understand the history and functions of artificial intelligence, the technologies used in autonomous vehicles and how they work. Knowing how autonomous trucks function will prepare workers for what’s coming and eliminates some of the fear surrounding them. It is a cornerstone in understanding how it will impact their work and is one of the first steps in understanding how they will need to adapt.A Brief History of Artificial IntelligenceWhen imaginingyou imagine artificial intelligence, scenes from the future and characters from mid-20th century science fiction may immediately come to mind.jump to the forefront of your mind: C-3PO and R2D2 from Star Wars, The Robot from Lost in Space that famously warned “Danger, Will Robinson!” or more nefariously, HAL 9000 from Kubrick’s 2001, A Space Odyssey. These fictional manifestations may seem like they can only exist in the distant future. However, there has been research and work performed on artificial intelligence for almost three quarters of a century. AI can be found everywhere, from our phones, to our cars, to our workplace.Artificial intelligence is the concept that a computer, through a series of complex scripts and algorithms, can “think” by taking in and processing large amounts of data. The possibility was first explored by British mathematician Alan Turing in his 1950 paper, Computing Machinery and Intelligence. He suggested that if humans can process information and use reason to make decisions and solve problems, a machine should be able to do the same thing. Turing’s vision was aimed at the future; at that point, computers were incredibly expensive and were only able to execute commands, not store them. However, as technology developed over time, that would change. The first artificial intelligence program, Logic Theorist made its debut in 1956. Over the following decades, AI grew and developed, thanks to dedicated research and some government funding. Now, with data being collected by every source imaginable, AI is beginning to have more practical applications.in its prime. It has uses everywhere, from banking and marketing, to entertainment, to autonomous vehicles. (Anyoha, 2019)Autonomous Trucks and How They Work. According to the California Department of Motor Vehicles, “Autonomous Technology” refers to any technology that is capabilecapable of driving a vehicle without the active physical control or monitoring by a human operator, while a vehicle in “Autonomous Mode” is in the status of operation where a combination of hardware and software, remote and/or on-board, performs the dynamic driving task, with or without a natural person actively supervising the autonomous technology’s performance of driving. (California Department of Motor Vehicles, 2019) In simpler terms, autonomous vehicles combine a series of complex technologies, such as interfacing with sensors and artificial intelligence, to operate without the input of a human operator.Automation is most effective when completing very repetitive tasks where a clear set of rules are in place. It seems driving, especially for long distances on the interstate, fits this bill quite well. Autonomous trucks will be rolled out in four waves. The first wave is currently in the works. Here, two drivers can create a “connection” between their trucks, not dissimilar to how you can connect your iPhone with a Bluetooth speaker. The drivers form a “platoon” and the trucks will maintain speed, stay in the same lane and remain a specific distance apart on the interstate to maximize fuel efficiency, until either driver inputs anything. When off the interstate, drivers operate individually. The second wave will likely take place between 2022 and 2025, according to McKinsey and Co. In this stage, the second driver is absent on the interstate between dedicated truck-stops. The driver in the lead truck provides all the necessary inputs, while the second, driverless truck will follow behind. When they reach a truck-stop close to their destination, another driver will climb aboard the second truck and take it the last few miles. Like the previous stage, the two trucks operate individually when off the interstate. In the third phase, the trucks operate under constrained autonomy. A driver will still bring the truck to a dedicated truck-stop, but now the trucks can operate entirely without a driver when on the interstate. When possible, trucks will “platoon” two or more vehicles when available. At this point, a driver will still have to climb aboard and take the truck the last few miles to its destination. Finally, the terminal stage of autonomous trucks will start sometime after 2027. Drivers are eliminated entirely, both on and off the interstate and will continue to platoon to maximize efficiency when -273050458533500-1886807875279Figure SEQ Figure \* ARABIC 1: A visual depiction of the four stages of autonomous trucks. Source: Chottani et al., 20180Figure SEQ Figure \* ARABIC 1: A visual depiction of the four stages of autonomous trucks. Source: Chottani et al., 2018available. (Chottani et al., 2018). With current technological advancements, it’s reasonable to assume we may be sharing the road with totally driverless vehicles within the next 15 to 20 years. Autonomous vehicles work by leveraging AI alongside imaging, positioning and communication technologies that vary from one level of automation to the next. In general, AT’s utilize cameras, GPS, lasers, sonar and radar systems to measure distance and get a sense of their surroundings. The AI acts as the vehicle’s “brain,” responding to the inputs that the sensors provide through advanced algorithms and predictive modeling. The AI then makes decisions based on those inputs, following the rules of the road, staying in the same lane, avoiding obstacles and keeping a reasonable distance between other vehicles. In the case of semi-autonomous vehicles, a driver may have to manually intervene if the AI faces any uncertainty, requiring the driver to always remain alert, even if they aren’t fully in control. (Union of Concerned Scientists, 2018)Applications to Supply Chain ManagementThe introduction of autonomous trucks will create many opportunities for firms to save costs. Platoon technology keeps multiple trucks a designated distance apart, improving aerodynamics, maximizing fuel efficiency and decreasing the cost of fuel. As automation improves, long-haul drivers may be phased out or transition into another role, decreasing labor costs. Drivers are also subject to federal regulation regarding how many hours they can work in a day. Currently, drivers can work for a period of 14 consecutive hours in a day, 11 of which can be spent driving. If more than eight hours have passed since the last time that they were off-duty, they are required to take a half hour break. At the end of that 14-hour period, drivers are required to spend 10 consecutive hours off-duty (United States Department of Transportation, 2015). Even when working in teams of two, there are still 2 hours of downtime per day where neither can drive. Autonomous trucks are not subject to the same regulations, so the time a product spends in transit is reduced. According to , the introduction of ATs will help the long-haul trucking industry grow 60% by 2024. However, these benefits are not only realized in the trucking industry. Improvements in the supply chain benefit nearly every firm. (, 2019) Reductions in lead-time lead to smaller inventories, which results in increased cost savings. The same lead-time reductions also mitigate a phenomenon known as the bullwhip effect, which has long been the bane of supply chain managers. ATs will ultimately change the strategies supply chain managers employ, but the return may be huge. Impact on Multi-Period Inventory ModelsFirms hold inventories for three reasons: to hedge against uncertain demand, to hedge against uncertain supply and supply delays and to take advantage of discounts for large order quantities or save on fixed ordering costs (advantages from making fewer large orders, rather than many small orders). Companies may have a lot of capital tied up in inventory, which has its inherent risks and ultimately impacts the bottom line. The more money that is tied up in inventory, the less there is to spend in other, more productive areas. Reducing inventories is a primary focus for supply chain managers. In a continuously monitored system, an order quantity is determined and a re-order point (ROP) is established. The ROP is the inventory level where a new order is placed. Optimal order quantity, notated as Q*, is the point where total cost (TC) is minimized and carrying cost (Cc) is equal to ordering cost (Co). Ordering costs, as the name implies, consist of the costs per order. This may include shipping and handling costs. Carrying costs are the cost per unit per unit time, multiplied by the quantity of units and the duration of time they are held. This may include costs for storage, insurance, pilferage, utilities, intra-company transportation, and the opportunity cost of capital. Annual Demand (D) is the amount of a product that consumers want to purchase. The formula for minimizing total cost (MinTC) is as follows:MinTC=CODQ*+CCQ*2From this we can derive:Q*=2CODCcBy decreasing the cost of labor and fuel, autonomous trucks will likely decrease the cost of ordering, Co. This, in turn, decreases the optimal order quantity, Q*. Both effectively lead to a decrease in total cost. In other words, the introduction of autonomous trucks will make it more economical for firms to place smaller orders, meaning less money is spent on inventory and profits increase. For example, imagine there’s a store that sells a product called a widget. The demand (D) for widgets is a constant 1 million per year. It costs the store $500 to have their supplier deliver an order (Co). Widgets can’t expire, but the cost of capital tied up in inventory amounts to 10% per year (Cc). The retailer’s optimal order quantity would be as follows:Q*=2*500*1,000,000.1=100,000We can use Q* to find the minimal total cost per order:MinTC=500 1,000,000100,000+.1100,0002=$10,000 In other words, to reach the minimal total cost of $10,000, the store should order 100,000 widgets per order. What would happen to the optimal order quantity if the supplier switched to a fleet of autonomous vehicles and the cost of ordering was reduced to $450?Q*=2*450*1,000,000.1 ≈94,868Again, we use that Q* to find minimal total cost:MinTC=4501,000,00094,868+.194,8682=$9,486.83In this case, if autonomous trucks reduce ordering costs by 10% from $500 to $450, the optimal order quantity is reduced by 5,132 widgets and the minimal total cost is reduced by $513.17.The re-order point is equal to the expected demand during lead time plus safety stock, the amount of inventory held in excess of expected demand. Average demand (?) is the average amount of the product sold. Lead-time (LT) is the amount of time between when the order is placed and when it is received. The standard deviation of demand (σ) represents how spread out the demand figures are. The target service level is the desired percentage of customers that won’t experience a stockout. The z-score (z) represents how many standard deviations a point is away from the mean and is calculated using the target service level on a standard normal table. The equation for finding the re-order point is as follows:ROP=μ*LT+zσLTAll other things being equal, if the target service level is above 50%, the reduction of lead times stemming from the use of autonomous trucks will decrease the re-order point, meaning that firms will place orders more often. Let’s return to the widget store example. If the annual demand is a constant 1,000,000 units, the average daily demand for widgets (?) is equal to 2739.73 units. Assume the standard deviation is 50 and the supplier’s lead time (LT) is 9 days. The store’s target service level is 98%, which results in a z-score (z) of about 2.05, according to a standard normal table. In this case, the re-order point would be: ROP=2739.73*9+2.05*50*9≈24,965Here, the store should place an order when they have 24,965 widgets left in stock. What would happen if the supplier began using autonomous trucks, resulting in lead time being reduced to 8 days? ROP=2739.73*8+2.05*50*8≈22,208By reducing lead time by 1 day, the reorder point was reduced by 2,757 units. Moral of the story: the introduction of ATs could make it more economical to hold fewer items in inventory and place orders more often. This means that supply chains are more responsive to changes in demand and firms have less capital in inventory, allowing them to spend elsewhere and improving profits. Little’s Law provides more evidence that Autonomous Trucks will reduce inventory levels. Little’s Law is as follows:Average Inventory=Lead Time*Flow Rate(Little, 1961) Assuming flow rate remains the same, the lead time reduction associated with autonomous vehicles will reduce a firm’s average inventory.205676521336000Mitigating the Bullwhip Effect24657052759710Figure 2: A graphical depiction of the bullwhip effect. Source: Lee, 19970Figure 2: A graphical depiction of the bullwhip effect. Source: Lee, 1997The bullwhip effect, first described by Jay Forrester in 1961, is a phenomenon in supply chain management where changes in demand order variabilities are amplified by demand fluctuations as you move up the supply chain. Put simply, a small variation in demand on the retail end of the chain leads to a fluctuation in order size by the retailer that results in larger order fluctuations at each ascending level of the supply chain (Lee, 1997). The bullwhip effect represents a problem for supply chain managers, as the distorted information ultimately leads to massive inefficiencies and wasted money. Minimizing the threat of the phenomenon is a primary goal for many managers. According to Lee, long lead times give firms an incentive to hold more safety stock. This ultimately results in greater fluctuations in order quantities over time when that stock is depleted than the demand data would indicate, exacerbating the bullwhip effect. However, the lead-time reduction associated with autonomous trucks would mitigate the effect, creating more accurate and responsive supply chains and saving firms money at every level of the supply chain. Industry DisruptionEven with these benefits, autonomous trucks are going to be disruptive and require a significant amount of planning and infrastructure development to ensure that the transition is smooth. Warehouses and distribution centers will have to operate 24 hours a day, 7 days a week to accommodate ATs; the trucks aren’t subject to human schedules anymore and may arrive at any time. Getting them loaded and departed as quickly as possible is crucial. Warehouses may also have to build loading docks and entrances that are compatible with ATs and would probably want to locate in areas with prime access to the interstate. Ports will face similar challenges as they need to keep up with ATs. Like the warehouses, they will also have to operate 24/7 and develop infrastructure to support the vehicles. Minimizing bottlenecks is also critical, so vehicles won’t get backed up. (Chottani et al., 2018) These developments require workers and firms to adapt. Constant operation may require a change in shift structure or hire more workers to meet new demands. In some career paths, such as the drivers themselves, their daily duties may change entirely. Truck drivers may see their roles change significantly, as they transition from long-haul driving to shorter deliveries and technological support. According to a Michigan State University study, once ATs are adopted on a wide scale, drivers must understand how to monitor the vehicles hardware, software and safety systems. Drivers must be able to diagnose issues and plan an immediate course of action when things go awry. Additionally, combined with the need to resolve the current shortage of long-haul drivers, current drivers may transition into localized logistics and delivery jobs. According to the American Trucking Association, 60,800 trucking jobs went unfilled in 2018, up almost 20% from 2017. If current trends hold, that number is predicted to grow to over 160,000 by 2028. (Costello et al, 2019) Whether there is a driver in the seat or not, a person is needed to attach the truck to the trailer, connect hoses from the truck to the trailer, perform routine maintenance checks or fill the truck with fuel. In the early stages of automation, a driver will have to remain in the truck and will have to drive it the last few miles to its destination. However, once more automation is introduced and trucks can operate independently on the interstate, the duties of a long-haul trucker will be radically different. Instead of driving, the “operators” are may have to keep an eye on the software, hardware and safety systems to make sure it’s running smoothly, as well as diagnose any immediate issues. They may also have to input data for trip scheduling and logistics and understand logistics software to make sure the planned route is effective (Yankelevich et al., 2018). As the need for driving skills decreases, the need for technological training and knowledge significantly increases. New in-service training and certification programs relevant to emerging technologies need to be developed to ensure the workforce is prepared for potential disruption.Case Study:While we know that the introduction of ATs may reduce lead-times, it is difficult to quantify the magnitude of these reductions. With no fully autonomous trucks on the road at this point, a simulated case is necessary to capture the difference in transport time between a human driver subject to federal regulations and an autonomous vehicle. The following case is very simplified and makes several assumptions. However, with all else being equal, it creates a model that demonstrates the scale of reductions solely due to the elimination of regulation.16189322761615Figure SEQ Figure \* ARABIC 3: Route Between New York City and Los Angeles Source: Google Maps, 20190Figure SEQ Figure \* ARABIC 3: Route Between New York City and Los Angeles Source: Google Maps, 2019161734512319000There is a long-distance trucking route between New York City, NY and Los Angeles, CA. The distance between the two points is 2,790 miles. The route involves taking I-80 W to I-76 W to I-70 W to I-15 S. Along those roads, the average speed for a truck is 57.7 MPH, 54.5 MPH, 56.8 MPH and 56.7MPH, respectively. (United States Department of Energy, 2011) This means that the average speed along all four roads is 56.425 MPH. The route is driven by a solo driver in a semi and the same truck fitted with autonomous driving technology. The two leave from the same point at the same time. The size of the load, traffic and road conditions are the same for both vehicles and no fuel stops are made. How much faster will the autonomous truck complete the trip?If driven straight through with no rest breaks or fuel stops, the trip would take 49.45 hours, or 2.06 days at that speed. However, truck drivers are subject to the following regulations regarding hours worked: (United States Department of Transportation, 2015)Regulations Regarding Hours Worked:How much can a driver work per day?The driver can work 14 consecutive hours in a day.How long can they drive for in a day?Within 14-hour working period, the driver can drive for 11 hours.How long do drivers need to rest for?After 14 hour working period, the driver must rest for 10 consecutive hours.Do they have to take breaks?If 8 consecutive hours have passed since the last off-duty period, the driver must take a 30-minute break.How long can they work per week?The driver is not allowed to be on duty more than 60 hours in 7 consecutive days, or 70 hours in 8 consecutive days.When do their weekly hours reset?60/70-hour limits "restart" after 34 consecutive hours off-duty.Figure SEQ Figure \* ARABIC 4: Summary of Regulations Regarding Hours Worked by Truck Drivers Source: United States Department of Transportation These regulations significantly improve driver health and safety. Tired, overworked drivers have a decreased ability to focus on the road, leading to more accidents, injuries and deaths. Regulations are necessary to protect the truck drivers and those who share the road with them. However, on long-distance trips, the regulations increase the amount of time it takes to reach the destination. The calculations to determine the difference in driving times due to regulations are summarized in the table below:Total Route Distance2,790 milesAverage Driving Speed57.7 mph+54.5 mph+56.8 mph+56.7 mph4 = 56.425 miles per hour.How many hours of nonstop driving will the trip take?2,790 miles56.425 miles per hour=49.446 hours or 2.06 days at minimum to complete the trip. This is how long it would take the autonomous truck without any breaks.How long would it take a human?49.45 hours11 hours of driving per day=4.495 days or 107.89 total hours to complete the trip.% Change4.495 days-2.06 days4.495 days=54.17%Figure SEQ Figure \* ARABIC 5: Summary of CalculationsThe difference in transportation times directly resulting in regulation to protect human drivers is significant. All other things, changing from a solo human driver to an autonomous truck on a trip from NYC to LA will reduce travel time by 54.17%. Time reductions, especially of this size, will result in more responsive, accurate supply chains, more efficient companies and higher profits. Ideally, cost savings will also be passed down to the consumer, resulting in cheaper goods for everyone. Automated Guided VehiclesIntroduction to Automated Guided Vehicles and WarehousingTrucks are not the only autonomous vehicles that are making a large impact on supply chain strategies. Warehouses and distribution centers are another key link in the supply chain. Warehouses are the center of a firm’s inventory and distribution operations. Collectively, American firms spent $1.64 trillion on warehousing and transportation costs in 2018. (Smith, 2019) According to the United States Department of Labor, over the past decade, the number of people employed in warehouses has near doubled from an average of 641,700 in 2009 to 1,188,500 between January and August 2019. (United States Bureau of Labor Statistics, 2019) With so much inventory moving in and out of warehouses, and so much money spent keeping them running, making warehouse operations more efficient and cost-effective are important opportunities to improve a firm’s profitability. Warehouses and their operations are evolving rapidly. For one, they’re getting bigger. The rise of e-commerce and demand for global distribution has led firms to build massive warehouses and distribution centers to effectively satisfy demand. As of October 2019, Trammel Crow Co. is building a $280 million, 3.8 million square foot distribution center in Liverpool, New York for a company that has yet to be revealed. Only one warehouse in the world surpasses that size, a 4.3 million square foot warehouse owned by Boeing in Everett, Washington. (Moriarty, 2019) With warehouses of that size handling such intense inventory loads, it is simply too much space and work to cover with human labor alone. One way that companies are looking to improve their warehousing operations is through the addition of autonomous warehouse vehicles, also known as Automated Guided Vehicles (AGVs). While autonomous trucks are currently emerging, Automated Guided Vehicles have been around for decades, and have made significant impacts. The first AGV, known as the Guide-O-Matic was introduced by Arthur Barrett and Barrett Electronics in 1954. The machine was a simple tow truck that followed a wire imbedded in the floor of a warehouse. Since then, AGVs have developed significantly more, especially their navigation systems. While early AGVs relied on hardware, such as wires or nails to find their paths, modern ones rely on systems like those seen in autonomous trucks. The effectiveness, flexibility, and efficiency of Automatic Guided Vehicles has led to a growth in popularity, and their wide adoption across many industries. (Olmi, 2011) Figure SEQ Figure \* ARABIC 6: Article on Barrett's Guide-O-Matic in the June 1958 edition of Popular Electronics Magazine. Source: Popular Electronics Magazine, 1958.The most modern AGVs take four main forms. The first type, known as trains, moves goodsmove goods between workers picking them and other areas. This can eliminate significant time workers spend moving from place to place in the warehouse. The second type is self-driving forklifts. Autonomous 3681095000forklifts use similar technology to autonomous trucks, such as laser navigation and camera setups to safely move around the warehouse floor, away from human workers. The third type monitors inventory. These machines go up and down the aisles of the warehouse scanning RFID tags on the materials in stock, keeping track of what is being used and what is in stock. This not only eliminates some of the need for physical inventory counts but can also help eliminate inefficient use of space and labor. The final type, drones, are being will eventually be used to move materials through the air around distribution centers. Like the 36810953938270Figure 7: An Autonomous Forklift Produced by Toyota. Source: .auFigure 7: An Autonomous Forklift Produced by Toyota. Source: .auinventory monitoring machines, they may also be equipped with RFID scanners to keep track of inventory. (Guillot, 2018)How Automated Guided Vehicles are Effective The immediate benefits of having autonomous warehouse vehicles are clear. First, while expensive initially, AGVs are far less expensive over their lifetime than the cost of the labor they supplement. According to the United States Department of Labor, the average wage of a nonsupervisory warehouse employee was $18.50 between January and July 2019. The average cost of an AGV is in the $100,000 to $150,000 price range, depending on the specifics of the vehicle. (Weber, 2015) This means that it would take between 5,405.4 and 8,108.1 hours of labor to recoup the cost of one AGV. On average, nonsupervisory warehouse employees worked 39.3 hours a week. (United States Bureau of Labor Statistics, 2019) Assuming an AGV is using a battery exchange strategy where the battery is replaced when it reaches a certain level, eliminating charging wait times, can operate 24 hours per day, seven days per week and is a substitute for three workers per day (one per eight-hour shift), it would take between about 45.8 and 68.8 weeks to pay off one AGV. After that breakeven point, firms will see a positive return on their investment in an AGV. In the short run, human workers are a less expensive option. Maintenance and repair costs are also significant and need to be accounted for when managers decide if an AGV is the right fit for their situation. However, in the long run, AGVs are generally more economical compared to the labor they replace. Labor reduction is not the only area where AGVs can provide significant benefits to warehouse operations. AGVs can potentially improve safety, reduce workers’ movement, and reduce workman’s compensation and insurance costs. Between 2011 and 2017, 614 people were killed in workplace forklift accidents, and more than 7,000 people are injured in forklift accidents per year. (United States Bureau of Labor Statistics, 2019) That means that someone is killed by a human-driven forklift in a workplace accident approximately once every 4 days. These incidents are tragic, unnecessary, and costly. Each year, approximately $135 million of immediate costs are incurred each year due to forklift accidents. (, 2019) AGVs can improve worker safety and save employers money. In order to protect human workers, AGVs are equipped with laser scanners designed to detect obstacles in the vehicle’s path with 100% accuracy, and a set of predetermined paths the AGV can travel is determined. (Cardarelli et al., 2017). AGVs are always paying attention, are always aware of their surroundings, and never stray from their programmed path. Every task is performed the same way every time. If human workers are properly trained to safely interact with AGVs, and stay out of their path, the amount of injuries involving vehicles may decrease significantly. There has not been a substantial amount of research quantifying the extent of these safety improvements, but cultural attitudes concerning AGVs and their production suggest that it is considerable. According to Fabien Bardinet, CEO of French AGV producer Baylo, human errors are frequent and normalized, but if an AGV malfunctions and makes an error, it is a much bigger deal. Scuff marks and dents from humans running warehouse vehicles into walls and shelves are a normal occurrence, but AGVs are held to a much higher standard. (Barthelemy, 2019) This higher standard ensures that AGV producers are creating consistent, safe vehicles, otherwise, the fearful attitudes surrounding AGVs would discourage managers from buying them entirely. One situation where AGVs may be most effective is when inventory is especially fragile and expensive or has low margins. At one large pharmaceutical production facility in New York State, warehouse workers frequently joke about the “Million Dollar Club.” Workers become members after they have damaged $1 million worth of inventory. It can be surprisingly easy to get to that point. In some instances, one pallet of materials can be worth tens to hundreds of thousands of dollars, so even a few mistakes add up against the bottom line. While not at all a source of pride, the club is a constant reminder of how costly human error can be. According to inventory management consultant Jon Schreibfeder of Effective Inventory Management, Inc. the equation to calculate additional sales needed to replace damaged goods is as follows:Additional Sales Needed= Value of Damaged MaterialAverage Gross Margin %(Schreibfeder, 2018) Therefore, if a company were to have a 20% average gross margin, and lost a pallet of materials worth $20,000, then they would require an additional $100,000 worth of sales to make up for that loss. If those margins were 10%, that number would double to $200,000. No matter what the average gross margins or the value of the damaged material are, one thing is clear: it takes more sales to make up for the loss of a damaged good than the value of the good itself. Eliminating human error is critical to decreasing costs of damaged goods. This is where AGVs can be beneficial. Automated systems are especially good at doing the same task repeatedly and will have little to no variation in how the task is performed. Humans may accidentally stray from the designated path, or not see obstacles. It’s these minor deviations that can be a major source of mistakes and costs. An AGV will follow the same path every time and its sensors will stop the vehicle if there is an obstacle. With these sources of potential human error eliminated, the amount of inventory items damaged should decrease, saving firms money. Like autonomous trucks, warehouse infrastructure must to adapt to meet the needs of AGVs. While the Guide-o-Matic and wires imbedded in floors may be a thing of the past, warehouses that utilize AGVs must make some accommodations for them. Simpler AGVs may require wires or magnetic tape to navigate the facility. Once these are laid down, they are difficult to move, meaning that if the warehouse were to be reorganized, it will take a significant amount of work to change the designated AGV path. More advanced AGVs with vision-guided navigation have the benefit of being more easily reprogrammed but require a lot of light to be able to “see.” AGVs with laser and lidar navigation should be programmed to avoid being in similar paths at the same time, because the systems may interfere with each other’s sensors, blinding the vehicle. (Gooley, 2016) It may be a challenge to implement AGVs in a warehouse if the existing equipment and processes aren’t complimentary. When considering an AGV, managers should account for the layout of the warehouse, and the potential for layout changes in the future.Issues and Potential ThreatsWhile the benefits of autonomous trucks and AGVs are bound to be significant, there are several issues and potential threats that could inhibit their adoption. These problems include both ethical and economic concerns. Understanding the issues and how we can approach them is a significant part of moving forward with introducing autonomous vehicles.In 2016, there were approximately 3,292,400 people employed as heavy and tractor trailer truck drivers, delivery drivers or driver/sales worker. (United States Department of Labor Statistics, 2019) According to a study performed by Frey and Osborne, 79% of tasks performed by heavy truck drivers and 69% of tasks performed by light truck and delivery drivers can be automated. (Frey and Osborne, 2017) Additionally, almost 1.2 million people are employed in warehouses. (United States Bureau of Labor Statistics, 2019) With so many people employed by these industries, there is serious cause for concern among many people. What will happen when all those jobs disappear? How will we retrain 4.5 million people? What are drivers and warehouse workers going to do for work after? While these fears are not totally unfounded, a 2018 report published by Michigan State University and Texas A&M found that job elimination will be more modest than many people think. Researchers estimate that long-haul trucking jobs will only decrease from 2.03 to 1.57 million and lightweight and delivery trucking jobs will fall from 1.5 to 1.12 million by 2028. (Yankelevich et al, 2018) They suggest that instead of eliminating jobs, autonomous trucks are going to change the demands of the workplace. Most trucking jobs probably aren’t going anywhere; it’s the functions of the job that are going to change. Drivers of the future must be more technologically inclined and be familiar with how the truck’s systems function. As hardware and software take over the primary driving functions, driving skills will take a backseat to technological skills. John Hitch argues in Industry Week that AGVs have similar effects on warehouse employment. According to Hitch, AGVs are doing the work that humans don’t want to do. Despite very low unemployment in the United States, firms are struggling to fill warehousing positions as humans become more selective with the jobs they perform. Some customers of Seegrid, an AGV manufacturer, state that they have 300% turnover in material handling drivers. (Hitch, 2019) With this trend in mind, it is less likely that AGVs will eliminate significant amounts of warehouse jobs, as they are largely replacing jobs that aren’t being filled. Training and certification methods need to be developed to prepare upcoming drivers for the future. People can be resistant to change sometimes, and drivers and managers alike might be hesitant to go all-in on autonomous trucks and AGVs because of their perceived disruptions. Ethical concerns have also been at the forefront of the conversation surrounding autonomous vehicles. Questions about what standards the AI will be programmed to and how it will make decisions when faced with dilemma situations have been asked by researchers, academics, governments and keyboard philosophers alike. If an autonomous truck was in a situation where it had to decide between protecting its driver and driving into a crowd of people or driving off a cliff and sparing the crowd, what should it chose? What if it came between hitting a young person vs. an old person? A human vs. an animal? Many people say that AI should be programmed to match human values. The issue is, which values do we use? According to an MIT study titled The Moral Machine Experiment, preferences in ideal AI behavior is largely dependent on region, with respondents in three distinct groups. Overall, respondents generally prioritized saving a group of people vs a single person, humans over animals and young people over old people. However, there was variations amongst people with similar backgrounds. The southern group, which includes France, Hungary and Latin America were much more likely to save young people. In the eastern group, containing most of Asia and the Middle East, where people traditionally have high value and respect for elders, respondents showed less preference for saving younger people. People in the western group, which includes most of Europe and the US, showed a strong preference for saving humans over animals, while the southern group was less likely to do so. (Awad et al, 2018)With different preferences around the globe, who decides what the ethics of AI should be based on? According to the German Ethics Commission on Autonomous Vehicles, dilemmas like those discussed in The Moral Machine Experiment should never happen. They state that decisions should never be made based on age, gender or appearance, although programming designed to reduce the number of injuries are justified. They also state that liability for damages caused by autonomous vehicles should be subject to the same principles as other product liabilities. This would ensure manufacturers continue to improve their systems over time, so damages become exceedingly rare. (Federal Ministry of Transport and Digital Infrastructure, 2018) However, will fear of liability disincentivize manufacturers from investing in developing autonomous vehicles? Unlikely. Even with the threat of liability looming, companies have continued to work on developing autonomous technology. If anything, it’s encouraging them to create the safest product possible. Situations that create ethical dilemmas, are far less likely with autonomous vehicles. Computers can’t get tired, distracted, or drunk. Accidents are going to become less common and these discussions less relevant.If there is one thing that will hold back the adoption of ATs and AGVs, it is fear and uncertainty. People are fearful of the economic implications. They worry about what will happen to transportation and warehousing jobs. They’re afraid that the composition of their careers is going to change. People are afraid that life and death decisions will be made by a computer. They may not trust the people behind the curtain that program the computer. A survey conducted by AAA found that 71% of Americans are afraid of riding in an autonomous vehicle. That number is 8% higher than it was in 2017. These feelings are justified, as accidents involving autonomous vehicles, including a deadly accident in March 2018, receive widespread media coverage. (Novak, 2018) However, the leading cause of these accidents? People. A survey of car accidents in California found that out of the 62 incidents involving moving autonomous vehicles, all but one was the fault of human drivers. (Reisinger, 2018) It’s these valid concerns that may ultimately slow down the introduction, but technology will continue to improve and eventually these worries will be quelled. Conclusion The introduction of autonomous trucks will create several benefits and new opportunities for drivers. In a route between New York and Los Angeles, an autonomous truck can complete the 2,790-mile journey up to 54.17% faster than its human-driven counterpart. The reduction in lead time that will come with their introduction makes it more economical for managers to place smaller orders more frequently, decreasing the amount of capital spent on inventories and positively impacting the bottom line. More efficient, reactive supply chains will be created, which may ultimately pass lower prices to consumers. However, the introduction of ATs does have its challenges. Autonomous trucks will be disruptive; they will change the way many people work, and some may be resistant to that change. There are also several ethical and economic concerns that arise.AGVs are making significant advances in warehousing operations. The most obvious benefit is their labor supplementation. When accounting for the labor they supplement, AGVs pay themselves off between 45.8 and 68.8 weeks. Additionally, advanced laser-guided navigation systems allow AGVs to “see” everything around them and follow the same path repeatedly. This eliminates variation and human error, increasing safety, and reducing worker injuries and damaged goods. Workplace injuries are a significant cost for firms, with forklift accidents alone causing $135 million in immediate costs annually. Damaged goods are also a significant threat for firms, as it typically takes significantly more sales than the nominal value of the damaged good to recoup costs. However, AGVs require proper infrastructure to be implemented successfully. More basic AGVs require wires or magnetic tape to be installed, which make it difficult to reprogram them if the warehouse is rearranged. Advanced AGVs with vision, laser, and lidar guided navigation systems are easier to reprogram, but each come with their own drawbacks. Vision-guided systems require large amounts of light to guide themselves, while other systems may interfere with each other if they get too close together. Warehouses must be designed with AGVs as part of their operating strategy. Trying to shoehorn AGVs into preexisting warehouse facilities will often result in a poor fit, resulting in negative performance.The fact is, autonomous vehicles haven’t been fully rolled out yet. It is impossible to know their true impact until they are available on a broad scale and more data on the subject exist. When that point is reached, further research with the available information will need to be done. Two major questions cannot be answered until this research is completed: will the true total cost of ATs outweigh the savings on inventory and labor they create? Will the reduction in transportation jobs reduce demand enough to counter any benefit from cost savings?Citations“About the Warehousing and Storage Subsector.” Bureau of Labor Statistics, U.S. Bureau of Labor Statistics, 18 Oct. 2019, , Rockwell. “The History of Artificial Intelligence.”?Science in the News, 3 Apr. 2019, sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/. “Autonomous Vehicle Ecosystem: Technology, Components, and Market 2019 – 2025.”?Market , Academic, 23 Jan. 2019, advantage.Product?pid=C745E1D6-1B87-4232-AAB6-2AF99A19F65C&view=SWOTAnalysis.Awad, Edmond, et al. “The Moral Machine Experiment.”?Nature, vol. 563, no. 7729, 24 Nov. 2018, pp. 59–64., doi:10.1038/s41586-018-0637-6.Barthelemy, Laurent. “Automated Forklifts Elevate Firms' Profit Hopes.”?, , 3 July 2019, of Labor Statistics, U.S. Department of Labor,?Occupational Outlook Handbook, Delivery Truck Drivers and Driver/Sales Workers,?12 Apr. 2019, Department of Motor Vehicles. “Key Autonomous Vehicle Definitions.”?Key Autonomous Vehicle Definitions, 15 Apr. 2019, dmv.portal/dmv/detail/vr/autonomous/definitions.Cardarelli, E., Digani, V., Sabattini, L., Secchi, C., & Fantuzzi, C. (2017). Cooperative cloud robotics architecture for the coordination of multi-AGV systems in industrial warehouses.?Mechatronics,?45, 1–13. doi: 10.1016/j.mechatronics.2017.04.005Clements, Lewis M., and Kara M. Kockelman. "Economic Effects of Automated Vehicles."Transportation Research Record, vol. 2606, no. 1, 2017, pp. 106-114.Chottani, Aisha, et al. “Distraction or Disruption? Autonomous Trucks Gain Ground in US Logistics.”?McKinsey & Company, industries/travel-transport-and-logistics/our-insights/distraction-or-disruption-autonomous-trucks-gain-ground-in-us-logistics.Costello, Bob, and Alan Karikhoff.?Truck Driver Shortage Analysis 2019. American Trucking Association, 2019,?Truck Driver Shortage Analysis 2019, Docs/News and Information/Reports Trends and Statistics/ATAs Driver Shortage Report 2019 with cover.pdf.“Employment, Hours, and Earnings from the Current Employment Statistics Survey (National).” Bureau of Labor Statistics, U.S. Bureau of Labor Statistics, Oct. 2019, . “Fact #671: April 18, 2011 Average Truck Speeds.”?, eere/vehicles/fact-671-april-18-2011-average-truck-speeds.“Fact Sheet | Occupational Injuries, Illnesses, and Fatalities Involving Forklifts | June 2019.”?Bureau of Labor Statistics, U.S. Bureau of Labor Statistics, 10 June 2019, .“Forklift Free: Driving Safety and Efficiency.”?, Seegrid, , Carl B., and Michael A. Osborne. "The Future of Employment: How Susceptible are Jobs to Computerisation?"?Technological Forecasting and Social Change, vol. 114, no. January 2017, pp. 254;280; -280Germany, Federal Ministry of Transport and Digital Infrastructure, Ethics Commission, et al. “Bmvi.de.”?Bmvi.de, Federal Ministry of Transport and Digital Infrastructure, June 2017. bmvi.de/SharedDocs/EN/publications/report-ethics-commission.pdf?__blob=publicationFile.Google Maps. “New York City to Los Angeles, CA.”?Google Maps, Google, 6 May 2019, maps/dir/New York City/Los Angeles, CA/@38.390031,-97.3109573,3.77z/data=!4m19!4m18!1m10!1m1!1s0x89c24fa5d33f083b:0xc80b8f06e177fe62!2m2!1d-74.0059728!2d40.7127753!3m4!1m2!1d-104.219293!2d40.2146073!3s0x876dcdbf3d2309f7:0x1d717c02687328e8!1m5!1m1!1s0x80c2c75ddc27da13:0xe22fdf6f254608f4!2m2!1d-118.2436849!2d34.0522342!3e0.Gooley, T. (2016, October 26). AGVs pioneer new paths in the warehouse. Retrieved from , Craig. “4 Types of Autonomous Mobile Robots, and Their Warehouse Use Cases.”?, Supply Chain Dive, 7 Aug. 2018, , John. "Reconciling Robot-Induced Anxiety and Admiration."?Industry Week, 2019. ProQuest, , Hau L., V. Padmanabhan, and Seungjin Whang. "The Bullwhip Effect in Supply Chains." Sloan Management Review, vol. 38, no. 3, 1997, pp. 93.Little, John D. C. “A Proof for the Queuing Formula: L = ΛW.”?Operations Research, vol. 9, no. 3, 1 June 1961.?PubsOnline, , Rick. “Distribution Center near Liverpool Would Be 2nd Biggest in World; Is It Amazon?”?, 5 Sept. 2019, .“‘No Hands’ Train.”?Popular Electronics, June 1958, pp. 58–58, , Matt. “71 Percent of Americans Still Don't Trust Autonomous Cars According to New Survey.”?Gizmodo, Gizmodo, 14 Mar. 2019, 71-percent-of-americans-still-dont-trust-autonomous-car-1833284527.Olmi, Roberto?(2011)?Traffic Management of Automated Guided Vehicles in Flexible Manufacturing Systems.?PhD Thesis?, Università degli Studi di Ferrara.Reisinger, Don. “Humans-Not Technology-Are the Leading Cause of Self-Driving Car Accidents in California.”?Fortune, Fortune, 29 Aug. 2018, 2018/08/29/self-driving-car-accidents/.Rouse, Margaret, and Diane Daniel. “What Is Supply Chain Management (SCM)? - Definition from .”?SearchERP, Feb. 2018, searcherp.definition/supply-chain-management-SCM“Self-Driving Cars Explained.”?Union of Concerned Scientists, 28 Feb. 2018, clean-vehicles/how-self-driving-cars-workSchreibfeder, Jon. “The Cost of Bad Inventory Control.”?Effective Inventory Management, 15 Apr. 2018, , Jennifer. “Logistics Spending Jumped 11.4% on Strong Economic Growth.”?The Wall Street Journal, Dow Jones & Company, 18 June 2019, .“Toyota Automated Guided Driverless Forklifts Offer Advanced, Flexible Solutions for More Efficient, Cost-Effective Goods Handling. Advantages of Automated Guided Vehicles (AGVs) Include Easy Tracking of Goods, Just-in-Time Picking, Less Damage and Fewer Operator Hours.”?.au, Toyota Material Handling, States, Congress, “Interstate Truck Driver’s Guide to Hours of Service.”?Interstate Truck Driver’s Guide to Hours of Service, FMCSA, 2015. fmcsa.sites/fmcsa.files/docs/Drivers Guide to HOS 2015_508.pdf.Wamsley, Laurel. “Should Self-Driving Cars Have Ethics?”?NPR, NPR, 26 Oct. 2018, 2018/10/26/660775910/should-self-driving-cars-have-ethics.Weber, Austin. “A New Generation of AGVs Are Appealing to Small- and Midsize Manufacturers.”?, Assembly, 20 Apr. 2015, , Clifford. "Back to the Future to Improve U.S. Transportation." Brookings Institution Press, 2010.Yankelevich, Aleksandr, et al.?Preparing the Workforce for Automated Vehicles. American Center for Mobility, 2018,?Preparing the Workforce for Automated Vehicles, comartsci.msu.edu/sites/default/files/documents/MSU-TTI-Preparing-Workforce-for-AVs-and-Truck-Platooning-Reports .pdf. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download