Autonomous systems projects are increasingly relevant to industrial automation, where sensing, control, robotics and decision-making must work together in demanding environments. The United States Naval Academy’s 2018 Senior Capstone Projects booklet documents a broad set of student designs that illustrate these engineering challenges. The work ranged from unmanned aircraft and underwater vehicles to inspection robots, control algorithms and energy-management systems.
The projects were presented as academic design and research efforts, not as validated industrial products. Even so, their technical themes provide a useful way to examine the problems automation engineers face when moving from an individual component to a coordinated system: defining operating conditions, integrating hardware and software, and planning for testing and maintenance.
Autonomous systems projects connect sensing to action
Many projects focused on turning sensor inputs into useful actions. One team proposed an unmanned system to locate, identify and map corrosion inside an enclosed structure. Another explored computer vision to detect moving targets on the water while compensating for a camera platform’s motion. A separate project investigated acoustic localization of unmanned aerial vehicles using a microphone array and time-difference-of-arrival methods.
These examples share a basic automation pattern: collect data, interpret it, and use the result to guide an operation. In an industrial setting, that pattern can appear in machine vision inspection, mobile robotics or condition monitoring. The project descriptions also make clear that sensing alone is not enough. A system must account for the environment in which measurements are taken—for example, motion that affects image stability or the need to navigate within an enclosed space.
For system integrators, this reinforces the importance of considering the full measurement chain. Sensor placement, operating conditions, data processing and the intended response all shape whether a system can perform its assigned task. A design that identifies a condition but cannot communicate it clearly or act on it safely is incomplete.
Robotics depends on integration and operating constraints
The program included several projects that combined mechanical design, control and mission requirements. An underwater robotic arm was proposed for inspection and maintenance of naval vessels. A corrosion-mapping robot was designed to move within an enclosed structure. Other teams worked on autonomous vehicles, including a conversion kit intended to enable a range of vehicles to operate autonomously.
These projects highlight a familiar challenge in industrial automation: a robot’s performance depends on more than its motion controller. The physical environment, payload, power source, access constraints and task requirements must be considered together. For an inspection robot, for instance, mobility and sensing must support the same objective: reaching relevant areas and identifying damage. For a vehicle autonomy kit, the design brief centers on adapting a system to different platforms rather than building a single-purpose machine from scratch.
The source descriptions do not report commercial deployment or production performance. Their value here is in showing how project requirements shape system architecture. Automation teams can use a similar approach when defining a project: specify the operating environment, the task, interfaces with existing equipment and the conditions under which the system must be tested.
Control strategies range from devices to swarms
Control research in the capstone program covered both individual machines and groups of autonomous agents. Projects addressed drone swarm coordination, autonomous landing of a fixed-wing aircraft on a moving platform, and control of a flexible underwater vehicle in simulation. A Trident Scholar project also examined persistent target tracking by multiple autonomous vehicles, including limited endurance, recharge requirements and coordination of operating modes.
These topics connect directly to industrial questions about distributed control and coordinated automation. When multiple machines share a task, the challenge is not simply to control each unit independently. Teams must also decide how responsibilities are assigned and how changing conditions affect the overall plan. The project descriptions show that endurance, sensing, communication and vehicle behavior can all be part of that problem.
Other work focused on control at a smaller scale. A smart-energy control system was proposed to sense measurable conditions such as lighting and temperature, then determine actions intended to minimize energy use without changing a room’s infrastructure. A separate project explored model predictive control to determine treatment for a virtual diabetic patient. Although their applications differ, both descriptions involve using system state and a control method to select an appropriate response.
Testing and maintainability belong in the design brief
Several projects explicitly included testing or validation in their objectives. A nuclear engineering team proposed physical tests of acoustic sensors on steel samples to investigate monitoring for stress corrosion cracking. Other teams planned experimental verification of control concepts using scale vehicles, models or simulation. One aircraft project was designed for field repair using materials available in the operating environment, while a separate project set a defined cost ceiling for an unmanned aircraft system.
These details point to practical questions for automation project planning:
- What conditions must the system tolerate, and how will they be represented in testing?
- How will sensor readings or control behavior be checked against a model or reference?
- Can equipment be inspected, repaired or adapted where it will operate?
- Which design constraints—such as space, power, cost or endurance—are fixed requirements?
Making these questions explicit early helps connect design decisions to operational needs. It also makes project reviews more concrete: teams can assess whether the proposed test plan addresses the conditions that matter, rather than treating testing as a final step.
What automation teams can take from the projects
The Naval Academy’s 2018 capstone portfolio spans engineering, mathematics, science and other disciplines. Its autonomous systems projects demonstrate a range of approaches to sensing, robotics, control and system integration, while also showing how tightly technical choices are linked to mission requirements.
For industrial automation engineers and plant leaders, the practical takeaway is to define the operating problem before selecting a technology. Clarify the environment, interfaces, performance requirements and maintenance expectations; then design testing around those conditions. Whether a project involves an inspection robot, a control algorithm or coordinated autonomous equipment, reliable automation depends on treating the complete system—not just its individual components—as the unit of design.