February 09, 2022 by Jeff Chu In a recent call with a customer, I was told "we're looking to build a new facility and investigating installing a tow-line to index our build carts down the assembly line. We really like the flexibility and scalability of an AGV or AMR, but we think a tow line may have higher ROI". On the surface, this customer’s initial inclinations seemed plausible. The assembly line under evaluation was 20 build stations, moving build carts that were designed for up to 4000 lbs each. As we dug into the application, however, we quickly realized that it was not necessarily true that tow lines are vastly more cost efficient than AGVs. To start – an initial question we reviewed is the takt time of the process. How much time is allocated to each workstation before the line must index forward? The customer’s process was set to 46 minutes. When assessing a process with long takt times for AGV vs tow line conveyance, there are two key considerations: If AGVs were to be used, is it necessary for each build cart to have its own dedicated AGV? If the process takt is long, how well balanced is the work content across the multiple stations? With more time in station, AGVs have more time to move between stations. It is therefore critical to understand if the AGV must be dedicated to each build cart or if the carts and AGVs can be separate entities. In many instances, there is no technical need for the AGV to be permanently connected to its companion frame cart that sits on top of it. In these scenarios, tunneling AGVs are a feasible solution that can lead to cost savings. This simulation highlights how only a small number of AGVs can be used to index an entire assembly line. As the name implies, a tunneling AGV is a vehicle that tunnels underneath a build cart and engages via a pin mechanism. While the pin is engaged, the cart is coupled to the AGV and moves to where the AGV takes it. Once the AGV positions the carts in their next stop, the AGV disengages and is free to go move something else. Longer takt also implies that a high volume of work content is likely being completed within each station. Given the complexity of products assembled today, it’s highly likely that the line balance is not perfect. Our second question above can be answered by observation. Are some operators complete their tasks before others and then waiting around? Is there a significant amount of idle time in some stations and not others? When considering the ROI for an AGV project in lieu of tow lines, it is critical to consider the flexibility that AGVs offer. In tow-line conveyance, if the work-content is not perfectly balanced, variability among workstation pacing often results in idle workers waiting for the slowest station (the Herbie) to catch up. In an AGV-based assembly process, each work station is de-coupled from the one prior. With buffer positions, this allows operators moving faster to call the AGV to move their unit down the line and start working on the next, even if somewhere downstream is still working – ultimately helping smooth out intra-station pacing variations. In the case of our customer looking to move a 20-station line every 46 minutes, there was huge opportunity for using only three AGVs instead of 20 to move all of the carts forward in the allotted time, and at a price comparable to that of a tow line. While I recognize this is just one anecdote with specific process design elements, the analysis shows that when considering an investment in a line conveyance technology, it is important to challenge mental models that each station needs its own AGV. With some creativity in cart design and thoughtful process engineering, AGV fleets can have high ROI with lots of added flexibility to your manufacturing operation. About the author Jeff Chu leads Factory of the Future Consulting at Eckhart, an Industry 4.0 solutions provider that supports some of the largest manufacturing operations in the world. Jeff has consulted for private equity and Fortune 500 clients across multiple industries and end markets, on topics including robotics, factory flow, and operations management. Jeff is a graduate of the University of Michigan and has an MBA and MS Engineering from MIT.
