Mind-controlled drones with Web AI & TensorFlow

Key Concepts

• Black Swan: The name of the company and its BCI (Brain-Computer Interface) technology.
• BCI (Brain-Computer Interface): A system that allows direct communication pathways between the brain and an external device.
• EEG (Electroencephalography): A neurophysiological monitoring method that records electrical activity of the brain.
• Swarm Control: The ability to control multiple drones simultaneously.
• Cognitive Load: The amount of mental effort required to perform a task.
• SDKs (Software Development Kits): A set of tools and libraries that allow developers to create applications for a specific platform.
• AI (Artificial Intelligence): Machine learning algorithms used to interpret brain signals and translate them into commands.
• TensorFlow: An open-source machine learning framework used for training AI models.

Black Swan: Revolutionizing Human-Machine Interaction with BCI

This presentation introduces Black Swan, a company pioneering a new era of human-machine interaction through Brain-Computer Interface (BCI) technology. The core innovation lies in enabling users to control devices, specifically drones, using only their brain and body signals, moving beyond the limitations of current interfaces like smartphones and watches.

The Need for a New Interface

The current day-to-day interfaces, such as phones, watches, and glasses, are acknowledged as functional but are deemed slow and limited in their command capabilities. Black Swan argues for a future where these devices, and indeed our environment, can be controlled solely by the brain. This vision is driven by several key issues with existing control systems:

• Slow and Limited Control: Current systems are not as fast or robust as desired for seamless interaction.
• High Cognitive Load in Multi-Drone Management: Traditional joystick systems are inefficient for managing multiple drones, as they typically allow control of only one at a time. This leads to significant mental effort for the operator.
• Lack of Intuitive Human-Machine Interface: The current need for extensive training to operate specific hardware is a barrier to widespread adoption. Black Swan aims for a "wear and use" experience.

Black Swan's BCI Solution: Oneperson Swarm Control

Black Swan's technology offers "Oneperson Swarm Control," a system that empowers any individual to command multiple drones using only brain and body signals. The current BCI headset can detect and interpret five distinct signals:

• Left and Right hand movements
• Head movements
• Blinking (identified as one of the loudest signals)
• Focus and defocus

These five commands are sufficient to operate their proprietary mind-controlled drone race, demonstrating the accessibility of advanced field robotics to non-experts. The technology is designed to be learnable in a short period, with the exception of individuals with severe ADHD.

Applications and Future Possibilities

The potential applications of Black Swan's BCI technology are vast and span multiple sectors:

• Esports: Enabling new forms of competitive gaming.
• Healthcare: Potentially aiding individuals with mobility impairments.
• Law Enforcement and Defense: Offering advanced control mechanisms for tactical operations.
• Real-time Human-Machine Teamwork: Facilitating seamless integration and collaboration between humans and machines.

Developer-Friendly Ecosystem

Black Swan is committed to making its technology developer-friendly. They are offering an operating system for BCI devices, allowing developers to program and customize its functionality. This includes the ability to control robotic hands or other programmable robots by connecting them to the BCI system. The company provides all necessary SDKs to facilitate this integration.

How the Technology Works

The Black Swan BCI system operates through a multi-stage process:

1. EEG Signal Acquisition: The BCI headset captures EEG signals from the user's brain. The current prototype uses four channels, with an eight-channel version in assembly. For comparison, healthcare facilities typically use 32 to 64 channels for brain scanning.
2. Data Transformation: The raw EEG data, received as a series of data sets in a command line, is transformed into actionable commands.
3. AI Interpretation: A layer of AI, trained using TensorFlow, interprets these transformed signals. This AI has pre-built commands and is being further developed to interpret user "thoughts" (e.g., up and down movements).
4. Command Transmission: The interpreted commands are then sent to the target device, such as a drone.

The company is actively engaging the public in its AI research, encouraging contributions to train their AI core models.

Live Demonstration and Pre-Orders

A live hardware demonstration showcased the BCI system in action. The process involves a brief calibration period where the user follows on-screen commands. The presenter demonstrated control by focusing and defocusing, which translated to drone movements. While safety protocols limited full flight in the demo, attendees were invited to book private demos for a more extensive experience. The system also includes safety features like automatic landing if a safe space is not detected, as seen with the DJI drone's behavior.

Call to Action and Future Outlook

Black Swan is currently open for pre-orders of their BCI headset, which features proprietary AI models and adjustable EEG sensors. A discount code is available for a limited time. Interested individuals can pre-order, book a demo, or even participate in an upcoming mind-controlled drone race on November 14th.

The company is also actively seeking investment and connections. They are open to fundraising and encourage interested investors or those with relevant connections to reach out. The presenter can be contacted via Instagram and Twitter (@Ogdar) or email.

Conclusion

Black Swan is positioned to be a pioneering force in the BCI space, offering a user-friendly, powerful, and accessible interface for controlling complex systems. Their focus on swarm control, developer enablement, and continuous AI improvement signals a significant step towards a future of seamless human-machine collaboration.