Optimizing Autonomous Mobile Robots: Seegrid’s Journey with Simio Simulation

Introduction

In the rapidly evolving landscape of material handling, Seegrid has established itself as a pioneer in autonomous mobile robot (AMR) technology. Supporting more than 50 global brands, Seegrid’s AMRs have transformed manufacturing and distribution facilities by automating up to 80% of non-conveyed material moves and reducing inventory requirements by up to 30%. However, as customer workflows grew increasingly complex, Seegrid’s Application Engineering team recognized the need for more sophisticated tools to design and validate their solutions. This case study explores how Seegrid partnered with Simio to implement discrete event simulation, enabling them to optimize AMR fleet operations, reduce congestion, and deliver more efficient solutions to their customers.

Figure 1: Seegrid Lift CR1 AMR in action.

Company Background

Founded in 2003 by Carnegie Mellon graduate Hans Moravec, Seegrid has deep roots in Pittsburgh, Pennsylvania. The company specializes in autonomous mobile robots that automate palletized material handling processes across manufacturing, logistics, and warehousing environments. Seegrid’s product line includes the Lift CR1 (capable of handling loads up to 15 feet high and 4,000 pounds), the Lift RS1 (for low-lift applications up to 6 feet and 3,500 pounds), and the Tow Tractor (for long-distance material transport with a capacity up to 10,000 pounds).

What sets Seegrid apart is their infrastructure-free navigation system built on stereo vision and 3D mapping technologies. This innovative approach allows their AMRs to navigate complex environments without requiring physical infrastructure modifications.

“There’s a revolution happening in material handling,” explains David Griffin, Seegrid’s Chief Sales Officer, “and at the heart of this revolution is Seegrid. Supporting more than 50 global brands, we continue to deploy our trusted and proven AMRs in live customer environments, completely transforming the way these facilities operate. The future is autonomous.”

The Challenge

Limitations of Traditional Planning Methods

While Seegrid’s Application Engineering team continues to leverage a dependable Excel-based model to support solution design, they needed to expand their approach with simulation modeling to provide deeper insights and greater flexibility when planning for more complex workflows.This tried-and-true method involved creating routes in CAD and then extracting that data to combine with speed profiles using an Excel macro.

“Knowing that our product has grown more and we are seeing more complex workflows with our customers, we knew we needed a different tool in our toolbelt,” Sydney Schooley, Seegrid Application Engineering Manager, explains. “Discrete event simulation helps us better understand and visualize Seegrid AMR interactions in customer environments.”

Specific Challenges in AMR Fleet Operations

The Application Engineering team faced several critical challenges that required a more robust planning system:

1. Visualizing AMR Interactions: Understanding how multiple vehicles would interact at intersections and identifying potential congestion points was difficult with static models.
2. Optimizing Job Sequencing: In modern manufacturing environments, particularly those with single presentation of parts, timing the replacement of empty containers with full ones (known as “hot swapping” or “job sequencing”) required precise coordination between multiple Seegrid AMRs.
3. Handling Peak Volumes: Determining whether the same fleet size could handle both average and peak volumes throughout the year was challenging without dynamic simulation.
4. Testing “What-If” Scenarios: The team needed a way to experiment with different fleet sizes and configurations to find the optimal solution for each customer’s unique requirements.

“Single presentation of parts is a common workflow seen in manufacturing material handling,” Schooley notes. “With manufacturing floor space at a premium, and with many parts that need to be added to a product, you often see one container of that material in a single presentation of parts.”

This scenario creates a time-critical situation—when a container of parts is depleted, the clock starts ticking for a material handler to exchange the empty container for a full one. With hundreds of parts needed for complex assemblies like automobiles, this creates significant congestion challenges for both human material handlers and autonomous systems.

• Job Sequencing Modeling: Simio enabled the modeling of complex job dependencies, ensuring that one AMR would complete its task (e.g., removing an empty container) before another AMR would begin its related task (e.g., delivering a full container).
• Traffic Analysis: The simulation provided detailed analytics on intersection wait times, helping to identify bottlenecks and optimize routing.
• Experimentation Framework: Simio’s experiment capabilities allowed the team to test different AMR fleet sizes and configurations to find the optimal solution for each customer’s needs.
• Multiple AMR Type Support: The model could simulate different types of Seegrid AMRs (like the Lift CR1, Lift RS1, and Tow Tractor) operating simultaneously within the same facility.
• Power Management: The model has the ability to account for the charging of AMRs during production. This logic is similar to Seegrid’s Fleet Central software, where once an AMR reaches a minimum battery level, it will go to the charger and be unavailable to receive new jobs until a minimum charge time has been met.

Figure 3: Battery Data Chart with charge and depletion rates per battery type.

Implementation Process

The implementation of Simio simulation at Seegrid followed a structured approach:

• Model Development: Seegrid formed a “Tiger team” consisting of Sydney Schooley, Application Systems Engineer Sofia Panagopoulou , and Application Engineer Abby Perlee, to collaborate with Sneath from Simio on developing the initial model.
• CAD Integration: One of the early challenges was integrating CAD drawings of facility layouts into the simulation model. The team developed a Python script that extracted route networks from CAD files and converted them into a format that could be imported into Simio.
• Fleet Management Logic Implementation: Seegrid’s enterprise fleet management software logic, Fleet Central, was translated into the Simio model to enable the simulation to accurately represent how the AMRs would behave in complex, real-world environments.
• Template Creation: The team created a template model that could be easily adapted for different customer scenarios, streamlining the simulation process for future projects.
• Team Training: Once the initial model was developed, the entire Application Engineering team was trained on using the Simio simulation tool, with a focus on practical application to real customer scenarios.

Schooley explains. “Our Tiger team was responsible for collaborating with Simio to help develop the model. Throughout the process—once we felt the model was in a good spot—our expanded team of 9 became more involved.”

Results and Benefits

The implementation of Simio simulation has delivered significant benefits to Seegrid’s Application Engineering team and their customers:

Enhanced Solution Design

The simulation model has allowed Seegrid to design more efficient and effective AMR solutions for their customers. By visualizing complex interactions and identifying potential bottlenecks before implementation, the team can optimize routes, fleet sizes, and job sequencing to maximize efficiency.

“Does this solution over time see major traffic jams at intersection 10? Does this add significant time to our cycle time? How can I reduce this traffic jam?” Schooley asks. “This is more problem solving on the [Seegrid] Application Engineers to better understand the customer situation.” It is on the skilled team of Application Engineers to determine how the model can optimize the final automated solution presented to the customer.

Improved Customer Communication

The visual nature of the simulation has elevated Seegrid’s ability to communicate complex solutions to customers. Being able to show a 3D visualization of how the AMRs will operate within the customer’s facility helps build confidence, internally and externally, for the proposed solution.

“And that visual is very impactful for my team,” Schooley notes.

Optimized Fleet Sizing

The ability to experiment with different fleet sizes has helped Seegrid optimize the number of AMRs required for each customer application. This promotes customers to invest in the right number of vehicles to meet their unique needs without over-purchasing.

“We’ll compare how changing volumes affect AMR fleet count,” Schooley explains. “For instance, if the customer has an average volume throughout most of the year, but then they peak during holidays, can we still use the same number of vehicles to help them at peak time as well?”

Efficient Job Sequencing

The simulation model has been particularly valuable for optimizing job sequencing in modern industrial manufacturing environments. By confirming that empty containers are removed before full containers are delivered, Seegrid can minimize the time that operators are without parts, boosting production efficiency.

“A process flow of job sequencing in manufacturing might look similar to this: An operator calls a container [Exchange Location] to be replaced through an integrated button press,” Schooley describes. “That signal starts a job. We’ll call it ‘Empty-for-Full Container- Parts 1, 2, 3, 4’. Then the Seegrid Lift AMR #1 will pick up the empty container from the assembly line and take it to a location designated for empty containers [Empty Storage]. Then AMR #2 would pick up a full container from a staging area [Full Storage] and place it in that same position at the line [Exchange Area]. AMR #2 will not start its job until AMR #1 has fully completed its job.”