AI is revolutionizing the engineering field, and as a part of their 2026 designs, students will integrate AI into engineering domains such as robotics, manufacturing, transportation, power systems, predictive maintenance, energy optimization, IoT and embedded systems.
A good design project involves more than the application of a machine learning model. A good project will also incorporate problem definition, measurable criteria, and engineering constraints on the system once it has been constructed and the performance has been assessed.
The same concepts are present in the latest iteration of the ABET engineering criteria. U.S. students can consult the ABET 2026–2027 Engineering Criteria for more information.
The following is a compilation of 15 AI-centric senior design project concepts for 2026.
15 AI Senior Design Project Ideas
| # | Project Idea | Best-Fit Majors | Core Technology | Expected Prototype | Key Evaluation Metrics | Difficulty |
|---|---|---|---|---|---|---|
| 1 | Edge-AI Pedestrian Crosswalk Safety System | EE, ECE, Robotics | Computer Vision, Edge AI | Camera + warning system | Recall, false alarms, latency | Advanced |
| 2 | Predictive Maintenance and RUL System | ME, IE, EE, Aerospace | Time-Series ML | Sensor/data dashboard | F1-score, MAE, RUL error | Advanced |
| 3 | AI Smart-Grid Fault Detection | EE, Power, Energy | ML, Forecasting | Grid simulation | RMSE, fault recall | Advanced |
| 4 | Autonomous Indoor Delivery Rover | Robotics, Mechatronics, ECE | SLAM, Computer Vision | Mobile robot | Navigation success, collisions | Advanced |
| 5 | AI PPE and Hazard Detection | Civil, Industrial, ECE | Object Detection | Camera monitoring system | Precision, recall, mAP | Intermediate |
| 6 | Battery State-of-Health Predictor | EE, Automotive, ME | Time-Series AI | Analytics dashboard | SOH MAE, RUL error | Advanced |
| 7 | AI HVAC Energy Optimization | ME, Energy, Architectural | ML, Optimization | Digital twin | Energy savings, comfort error | Advanced |
| 8 | RL Traffic Signal Controller | Civil, Transportation, CE | Reinforcement Learning | Traffic simulation | Waiting time, queue length | Advanced |
| 9 | AI Water-Network Leak Detection | Civil, Environmental | Anomaly Detection | Network simulation | Detection rate, false alarms | Intermediate |
| 10 | AI Manufacturing Quality Inspection | Industrial, Manufacturing, ME | Computer Vision | Inspection station | Defect recall, inspection time | Intermediate |
| 11 | Assistive Navigation Wearable | Biomedical, EE, ECE | Vision, Sensor Fusion | Wearable prototype | Accuracy, latency | Advanced |
| 12 | AI Smart Irrigation System | Agricultural, Environmental, EE | IoT, ML | Sensor testbed | Water use, prediction error | Intermediate |
| 13 | RAG Engineering Maintenance Copilot | Computer, Software, Industrial | LLM, RAG | AI assistant | Retrieval accuracy, hallucination rate | Intermediate |
| 14 | AI Surrogate Model for Design Optimization | Aerospace, ME, Chemical | ML, Optimization | Simulation tool | RMSE, simulation speedup | Advanced |
| 15 | AI Process Fault Detection | Chemical, Industrial, Control | Anomaly Detection | Process monitoring | Fault recall, detection delay | Advanced |
1. Edge-AI Pedestrian Crosswalk Safety System
Build a system that uses cameras to identify pedestrians at crosswalks and uses that information to trigger an alert signal. Constructs can use OpenCV, vision models, Raspberry Pis, or Jetson devices to complete this project.
The system will be evaluated based on: pedestrian recall, false alarms, detection latency, low-light performance.
2. Predictive Maintenance and Remaining Useful Life
Use vibration, temperature, current and historical equipment data for predicting equipment faults or Remaining Useful Life.
NASA has an excellent dataset for engineering use and maintenance related to prognostics in its Prognostics Data Repository.
Random Forest, XGBoost, LSTM, CNN, or anomaly-detection methods are possible models.
3. AI Smart-Grid Fault Detection and Load Forecasting
Create an AI application for electricity or renewables demand forecasting or power system faults detection.
You can use Python, but you can also use MATLAB/Simulink. Also explore Fried Engineers’ following resource related to smart-grid load forecasting based on transformer-based federated learning.
4. Autonomous Indoor Delivery Rover
Design an autonomous indoor rover to navigate to defined delivery locations while avoiding obstacles.
Possible technologies include ROS2, Gazebo, LoDAR (Light Detection and Ranging), IMU, and OpenCV, along with control systems on either a Jetson or Raspberry Pi.
Assess performance through navigation success, collision rate, localization error, and travel time.
5. AI PPE and Hazard Detection System
Design a computer-vision system to detect hard hats, safety vests, and other safety equipment or to detect unsafe conditions by entering restricted zones.
The system will integrate a camera, AI detection, safety logic, safety event logging, and safety alerts and notifications.
6. Battery State of Health Predictor
Estimate battery State of Health (SoH) and Remaining Useful Life (RUL) from measurements of voltage, current, temperature, capacity, and charge cycles.
This is relevant for EVs, robots, and renewable energy and energy storage systems.
7. AI HVAC Energy Optimization
Generate a building or room model that balances temperature, occupancy, humidity, weather, historical energy consumption, and activity within the building or room to optimize the HVAC system.
Evaluate the impact of AI control on the energy consumption and comfort metrics, along with traditional control as the baseline.
8. Reinforcement Learning Based Traffic Light Controller
A Controller designed using Reinforcement learning can learn how to optimally determine the timing of traffic lights based on congestion levels and length of queues.
The impact of AI Control over traditional Fixed-time Control can be studied using the open-source SUMO traffic simulator.
9. AI for Leak Detection in Water Networks
AI can help identify leaks and abnormal operating conditions in water networks by analyzing pressure and flow data.
To develop a realistic simulation of water networks, EPANET software provided by the U.S. Environmental Protection Agency can be used.
10. Edge AI for Quality Inspections in Manufacturing
Design a system that uses computer vision to detect manufacturing defects such as scratches, missing parts, and assembly errors.
Measure defect recall, false rejection rate, inference speed, and inspection time.
11. AI-enabled Wearable Assistive Navigation
Design a computer vision and distance sensor based system that detects obstacles and provides feedback via audio or vibration.
The design should consider accessibility, privacy, response time, and system failure.
12. AI-based Intelligent Irrigation Control System
Design a system that predicts irrigation requirements by analyzing soil moisture, temperature, and humidity data.
Refer to the Smart Irrigation using IoT and ML by Fried Engineers.
New users can also refer our IoT Project Development Guide.
13. RAG-based Assistant for Engineering Maintenance
Design an AI assistant that provides answers to technical queries using engineering manuals and maintenance logs through Retrieval-Augmented Generation.
Analyze retrieval accuracy, citation accuracy, hallucination rate, and response latency.
The NIST AI Risk Management Framework is useful for Trustworthy AI design.
14. AI-based Surrogate Model for Engineering Design
Design a ML based model that can replace complex and time consuming simulations of aerodynamics, thermodynamics, structure, and processes.
Don’t forget to plug the surrogate model in your optimization algorithms.
Examine another similar Fried Engineers project involving the use of machine learning to optimize aerodynamic shapes.
15. AI Process Fault Detection System
Identify process anomalies using variables such as temperature, pressure, flow, level, or consumption of a process energy.
Techniques like PCA, Isolation Forest, Autoencoders, LSTM and certain hybrid models of ML and physics could be used.
How to Pick the Right AI Senior Design Project
Some critical questions are
- Are there defined problem statement and solution?
- How does AI affect the solution? Is it for the better?
- Is it possible to obtain the data you need?
- Can we define whether a solution is successful or not?
- Can your team develop a working prototype in the defined time and cost constraints?
Complexity does not guarantee quality.
A project with well synthesized parts is often more impressive than an over ambitions one that ends up incomplete.
Prior to the finalization of your Bill of Materials, hardware-based projects should utilize the Fried Engineers Engineering Component Selection Framework.
Frequently Asked Questions
What is the best AI senior design project for engineering students?
Predictive maintenance, robotics automation, algorithms within smart grids, manufacturing inspection, intelligent transportation systems, battery health prediction, and edge AI systems are some effective and measurable integration of AI and engineering projects.
Do senior design projects require hardware?
They definitely do not have to. Optimization of traffic, digital twins, RAG systems, intelligent forecasting of smart grids, surrogate modeling, and predictive maintenance can also be effective software – or simulation – based capstones.
Can engineering students use ChatGPT or other LLMs in senior design?
As long as this technology is approved by the University and AI is providing a valuable service, then yes. Generating RAG responses, citation checking, engineering document retrieval, hallucination detection, or the integration of other tools can be done at a much higher level of engineering than a standard chatbot.
How can an AI senior design project become more research oriented?
Add a clear research question. It can be the comparison of algorithms, the placement of edge vs. cloud inference, the impact of sensor noise, interpretation and ranking of AI explainability, modeling and placement of a physics based vs. ML approach, or the performance under different operating conditions.
Final Thoughts
While AI increases capability, engineering senior design projects should integrate AI to provide a creative and effective engineering solution rather than add AI for the sake of adding something new.
Describe a problem and articulate some goals with pragmatic measurement metrics. Determine what tools are necessary and build the system. Evaluate the system and outline both the positives and negatives of the system.
Check out Fried Engineers for more engineering project and research resources. Their website provides many research-paper-based project directions across various engineering fields.
Create something that is testable and defensible.
