Imagine a naval aviation squadron schedules officer spending all day working on many iterations of the next day’s flight schedule, only to find out from the maintenance team that two more jets are broken. Instead of having 14 jets for the schedule, now there are not enough aircraft. The schedules officer is exhausted but has to rewrite the schedule in a way that will minimize the events lost and maximize the training. This scenario is quite similar to a challenge commercial airlines commonly face with daily operations plans, with one key difference: Commercial airlines use artificial intelligence and machine learning to automate the writing and revising of flight schedules. The Navy uses the same process it has used for 100 years—stick a junior officer in a room to figure it out.
The Navy wastes thousands of man hours and millions of dollars each year perpetuating fully manual processes in lieu of automated solutions. Naval aviation would be a more tactically proficient and lethal force if it were to incorporate semiautomated scheduling software solutions. The community must transition to a computer-augmented flight schedule generator to remain a premier air force and retain the best aircrews in the world. Not only will such a move save the Navy a tremendous amount of money, but it also will reduce the resources and administrative overhead necessary to complete squadron training detachments. Implementing a semiautomated scheduling solution to the fleet replacement squadrons would be a good way to work out the bugs before scaling up a fine-tuned solution to all of naval aviation.
Training Detachment Flight Schedules
Every six weeks, one F/A-18 E/F Super Hornet fleet replacement squadron, that trains junior naval aviators how to fly and employ tactical weapon systems, goes on a two-week training detachment to Naval Air Station Key West, Florida. Preparations begin months in advance and involve coordination among internal (squadron maintenance, operations, training, and safety officers) and external (other squadrons vying for range time, Naval Air Station Key West base operations, and adversary support aircraft) contributors. Fleet Replacement Strike Fighter Squadron (VFA) 106, from Naval Air Station Oceana, Virginia, brings 18 aircraft (worth about $75 million each), 165 maintainers, and 40 aircrew (at a personnel cost of about $500,000 per detachment). While in Key West, the student class flies about 400 events to satisfy the requirements of the strike-fighter aviation syllabus.
Scheduling the daily flight operations of such a squadron is challenging and time consuming, and the schedules officer position is rotated to a different pilot each day to prevent burnout. Scheduling daily flight operations involves coordinating five major elements: replacement (student) aircrew, instructor aircrew, airspace, jets, and opposition air support. A common refrain heard throughout squadron operations offices is that a schedules officer must be able to “see the matrix” to be successful—thinking through all the possible schedule iterations, incorporating intangibles such as personalities, optimum skills, individual desires, weather, and many more. A great schedules officer can get close to a 90 percent solution, leaving 10 percent of potential missions on the table. This means that despite thousands of hours of flight time and years of experience, the best that can be expected is about 90 percent operational efficiency. A semiautomated software scheduling program could improve that percentage.
Old, Tedious Scheduling Software
The schedule template software the Navy uses is version 7.0 of the Sierra Hotel Aviation Readiness Program (SHARP), an out-of-date program that allows schedules officers to place names with aircraft, missions, and airspaces. However, it requires repeating the same actions many times to produce the schedule. For example, when a schedules officer enters names then goes to the next line on the program, the program should intuitively number the events—the first event is #1, the second #2, and so forth. This is not the case, however, and so for every event, the schedules officer must spend mindless hours repetitively inputting the same data.
This manpower-intensive process is why VFA-106 no longer uses SHARP while on detachment to Key West. Instead, VFA-106 automated this process with an Excel file it calls “Shell-It.” It has a few thousand lines of code and is used by the schedules officer to generate a flight schedule without the onerous and time-consuming requirements of SHARP. Schedules officers still must generate the most important parts of the flight schedule—replacement aircrew/instructor aircrew/mission/airspace/red air pairings—but they do not have to worry about interfacing with a slow and burdensome online program at a remote air station notorious for poor network connectivity. With Shell-It, the schedules officer still must “see the matrix,” because it is these distinct pairings that help support the goals of VFA-106, but he or she can devote more time to the pairings and less time working with SHARP. Shell-It increases efficiency.
Imagine if on every detachment, VFA-106 was able to complete students one day early. In a year, the time savings would result in an extra week of flight events that could be allocated elsewhere. Implemented across all of naval aviation, thousands of flight hours (at about $10,000 per hour) and man-hours would be saved.
Progress Toward a Semiautomated Solution
The Navy is not alone in its flight scheduling woes. While the Air Force still relies on mostly human labor to generate its training pairings, it recently recognized the limitations of this type of scheduling and has made significant strides to improve its process. Until last year the Air Force used a whiteboard, dry-erase markers, and extensive hand-made spreadsheets to plan daily tanker aircraft operations. This required multiple people working eight hours a day to generate the flight schedule. The Air Force and the Pentagon’s Defense Innovation Unit in Mountain View, California, signed a $2 million contract with a commercial company to develop a semiautomated solution, and after 90 days had a program that saved more than $1 million per day. The tanker scheduler office now has one person spending just four hours per day guiding the computer that crafts the schedule.
Similarly, VFA-106 is partnering with the U.S. Naval Academy’s Mathematics Department and operations-research team to design an automation program that allows the squadron schedules officer to generate a schedule while on detachment in less than an hour. This program, called “Shelly,” was recently tested in Key West and proved to beat the schedules officers’ solutions by at least a 10 percent margin. Shelly can write a more efficient flight schedule in 14 minutes than a four-man team of highly skilled fleet aviators can in eight hours.
What would it mean to a squadron operations officer to have the equivalent of ten extra days of training per year? What would it mean to naval aviation if it could qualify an extra pilot per year based on the savings? How many aviators leave the service because they become frustrated working too many hours doing mundane, repetitive tasks that can and should be automated? The worst that could happen from the Navy allocating funds toward the VFA-106 project and others like it would be that it gives some hope to junior officers that the Navy values their time and intellect. These junior officers want to focus on tactics but are forced to troubleshoot computers each day. Ultimately, they decide that spending ten minutes watching a software program fail to load is not what they want to do for the next 20 years.
Leaders must realize that budgetary decisions have to be pushed to the lowest possible level so the Navy can better innovate. Squadron flight scheduling can be partially automated. The Navy should hire a company to tackle this limited goal. Once successful, such an innovation will pave the way for greater automation across all of naval aviation and the Navy.