We Built an Ai Laser that Destroys Moving Targets!
Science & Technology
Introduction
Some months ago, we shared an experiment where we mounted a powerful laser assembly on a DJI gimbal alongside one of our Lumex video cameras. We aimed to create a system that uses real-time video feedback to enable the laser to track and destroy targets. While it was a fun project, we encountered several shortcomings and decided to significantly enhance the original system with some help from MIT’s student-run group, MITERS.
Collaborating with MIT
We reached out to MITERS to assist in upgrading our system, and we collaborated with Chris, a graduate student studying Robotics. The improvements we wanted required advanced technology skills that we didn’t possess, so this partnership was instrumental. Chris introduced us to various algorithms and control systems needed to effectively track targets with our enhanced laser system.
Upgrading the Laser Assembly
The original laser assembly consisted of a multitude of individual laser diodes capable of producing up to 600 watts of visible blue light. Our upgraded laser has 28 diodes, allowing it to generate over 1,600 watts of laser light—a remarkable increase in power. This increase allows us to achieve extraordinary intensity, making our laser one of the most powerful of its kind.
We conducted a series of tests to see what the new laser could do. After preparing the equipment, we engaged the laser on targets like aluminum oxide fire bricks and stainless steel sheets. The results demonstrated that we had the necessary power to melt through tough materials.
Addressing Challenges with Fluorescence
While testing, we found that the laser’s intensity caused phenomena like fluorescence, which made tracking targets unexpectedly challenging. The emitted light from targets was so bright that it often washed out the camera’s visibility. To address this challenge, we replaced the camera’s bandpass filter with a short-pass filter that could block unwanted fluorescence while still allowing essential light for tracking.
An upgrade to a monochrome camera increased sensitivity to blue light, and we added other lighting solutions to ensure the camera functioned effectively indoors.
Control System and Tracking
The tracking system utilized a gimbal powered by stepper motors, allowing us to adjust the laser direction with precision. The camera sends images to a computer, which processes them through an object detection network called YOLO. This allows the system to consistently identify and track objects—RC cars, for example—within its field of view.
We set up a loop whereby the camera detects the object’s position, and our control system calculates corrective adjustments to the laser's position, allowing it to accurately stay aimed at the targets.
Final Demonstrations and Tests
In several experiments, we tested the laser's tracking capabilities on moving targets, including balloons and RC cars. The results were thrilling, as our system seamlessly tracked and engaged the targets.
During our final demonstration, we engaged drones in flight as our laser targeted them. The power intensified, successfully disabling multiple drones while allowing us to observe the effects of our high-powered system in real-time.
Future Prospects
The project highlighted both the capabilities and the potential risks of laser technology. As lasers become more powerful and more affordable, it raises concerns about their potential misuse. However, these technologies can also be leveraged for beneficial applications, such as detecting chemicals in various industries.
We’re committed to continuing our research and development in this domain, recognizing both the ethical implications and the exciting possibilities ahead.
Keywords
AI Laser, MITERS, Laser Assembly, Real-Time Tracking, Object Detection, YOLO, Target Destruction, Robotics, Control Systems.
FAQ
Q: What is the power output of the upgraded laser?
A: The upgraded laser has the capability to produce over 1,600 watts of visible laser light.
Q: How does the system track moving targets?
A: The system uses a camera to take pictures of the environment and an object detection algorithm (YOLO) to track identified targets in real-time.
Q: What challenges did you face during testing?
A: One challenge was the intense fluorescence from the targets, which washed out the camera’s visibility and made tracking difficult.
Q: How do you ensure safety while testing the laser?
A: Strict safety measures are employed, including locking doors, using goggles, and implementing a kill switch in the system.
Q: What are the future applications for this technology?
A: Potential applications include advanced targeting systems, chemical detection in industrial processes, and environmental monitoring.
