Robotics & Automation

Robots and automated processes have been transforming operations in global companies for decades. LGO students are able to focus their studies, both in the classroom and in their internship project, on robotic solutions to business problems. The Aeronautics and Astronautics, Mechanical Engineering, and Electrical Engineering departments all offer robotics classes, and MIT’s cross-disciplinary institutes like CSAIL and the Media Lab all contribute to groundbreaking research on the topic. Many of our partner companies are also eager to leverage LGO knowledge into new robotics solutions.

Automated Assembly System Process

Ellen Ebner (LGO ’15)

Company: Boeing
Location: Charleston, SC

Problem: The Boeing Company’s Propulsion South Carolina (PSC) is responsible for design and assembly of the engine nacelle inlet for the 737MAX, supporting a vertical integration strategy that differs from recent airplane programs. Development of the new 737MAX inlet factory is a unique opportunity to build an automated assembly system to meet high production rates and reduce the variability in quality and safety related to extensive manual drilling.

Research showed that there had been acceptance testing and requirements verification has been inconsistent for similar automation projects throughout Boeing. Ellen was asked to develop a comprehensive buyoff plan to:

  • Create a functionally sound system in the Boeing environment
  • Avoid a schedule setback or cost impact due to non-value-added work during testing period or production
  • Increase consideration of requirements across functions
  • Leverage Boeing expertise in aerospace engineering requirements, lessons learned from past buyoff processes, and robotics vendor Kuka’s expertise in automation system and component testing methodologies
airplane manufacutirng internship
Ellen’s requirements traceability and verification matrix.

Approach: Ellen created a tool to trace how system requirements became testing activities.
She was able to develop a way to openly show system performance to teammates and leadership. She then developed a plan to enhance cross-discipline requirements dialogue through involving stakeholders early in the process, tracking participation, and managing multi-interest changes to the buyoff plan.

Impact: Stakeholders helped create the buyoff plan and have utilized the tools to trace how requirements are fulfilled. The plan also helps to define acceptance-testing practices and creates ownership for testing activities. The buyoff plan has helped lower the risk that a system will not perform on standards. Ellen’s acceptance testing schedule also avoids budget overrun associated with unspecified acceptance criteria and ownership.

Risk Analysis of Unmanned Aircraft Systems in National Airspace

Jacquelyn N. Mohl (LGO ’16)

Company: National Grid
Location: Boston, MA

Problem: National Grid provides safe, reliable electricity and gas service throughout the northeast United States. To maintain its assets, technicians need to inspect assets on both regular and emergency bases. Currently, National Grid utilizes helicopter inspections, which can be costly and schedule-constrained. They also use on-foot inspections, which can be dangerous or at times impossible. Thus, the company is looking to incorporate unmanned aircraft systems (UAS, or drones) into its operations. National Grid asked Jackee to provide a roadmap to incorporate UAS into National Grid’s operations. National Grid also wanted a comprehensive risk analysis for utility companies to guide the Federal Aviation Administration’s (FAA) imposed restrictions, seeking in effect wider use of UAS in utility operations.

mohl image
A probability model for UAS ground collisions.

Approach: Jackee identified the technical and business requirements for internal use cases, evaluated UAS technology, benchmarked with outside utility companies, developed safety and training procedures, and submitted an application for FAA approval. With this approval, National Grid can move forward with a pilot program to evaluate UAS technology and data, further refine operating procedures, and make decisions on implementation.

The FAA application process uncovered a large amount of restrictions the agency imposes on commercial entities using UAS. For example, the regulations would make it impossible to conduct post-storm damage assessment using drones. Jackee conducted a probability risk analysis. She then determined a level of safety (LOS) for utility applications based two failure modes that she identified by a hazard tree analysis: midair and ground collisions.

Jackee developed a ground collision model using UAS reliability, wind gust data, population density, time-use survey data, and the FAA-published LOS for aircraft in national airspace. She determined the minimum UAS mean time between failure (MTBF) required to meet the FAA LOS requirement. She then concluded that the UAS must have a minimum MTBF of 250 hours to operate within National Grid’s service territory. This MTBF measurement varies by town due to differences in wind speed and population density. Jackee developed a similar model for midair collision probability analysis.

Impact: Through this risk analysis, National Grid and other utility companies have additional data on the safety of operating UAS for their operations. National Grid can use these findings to potentially push back on the FAA’s restrictions and open up more use cases for utilities.

Velocity Inventory Management in Automated Warehousing

Jake Stowe (LGO ’16)

Company: Amazon (Kiva Robotics)
Location: Dallas, TX

Problem: The transition from labor-intensive to capital-intensive processes marks a major change for the warehousing industry. In the past, volume constraints depended heavily on staffing levels and managers could scale their operations to meet demand inexpensively. The advent of multi-agent warehousing automation platforms such as Amazon Robotics (formerly Kiva Systems) has fundamentally changed the physics of inventory management. Head count can no longer be used to strongly influence outbound volume, so managers are looking for other levers to optimize these new systems.

Stowe image
Amazon’s order-picking equipment created by Kiva Systems.

Approach: During his internship, Jake suggested, analyzed and optimized one potential lever: velocity stow.  Velocity stow is the sortation of SKUs into multiple classes based on their “velocity” (sales per unit time) and stowing these classes in an organized manner into the inventory field. To put it differently, spatial organization of SKUs is based on forecasted sales and historical sales data.

Jake constructed a theoretical model relating the velocity of SKUs stored on individual racks. He then simulated how different velocity stow strategies impact distributions of potential pick accumulation on racks and the mean mission times of bots (two metrics that are determinants of hardware and operator utilization of the system.)

Impact: The simulation demonstrated that a velocity sortation strategy increased the potential pick accumulation on the high-velocity racks. Amazon Robotics use Jake’s analysis to optimize how their facilities are organized.