About this project
OpenBCI is a low-cost, programmable, open-source EEG platform that gives anybody with a computer access to their brainwaves. Our vision is to realize the potential of the open-source movement to accelerate innovation in brain science through collaborative hardware and software development. Behind the many lines of code and circuit diagrams, OpenBCI has a growing community of scientists, engineers, designers, makers, and a whole bunch of other people who are interested in furthering our understanding of the brain. We feel that the biggest challenges in understanding what makes us who we are cannot be solved by a company, an institution, or even an entire field of science. Rather, we believe these discoveries will be made through an open forum of shared knowledge and concerted effort by people from many different disciplines. By donating to this Kickstarter you will be supporting our efforts to make the existing OpenBCI hardware ready for mass production and available to the general public. In addition, you will be joining the OpenBCI community where people of all ages, locations, and backgrounds can contribute to unlocking the mysteries of the human brain.
Brain-computer interfacing (BCI) is a relatively new field of science with a seemingly limitless range of applications. Medical grade BCIs are often used in assisting people with damage to their cognitive or sensory-motor functions, however, more and more we are seeing affordable BCIs emerge in neurotherapy applications that assist people with ADHD, anxiety, phobia, depression, and other common psychological ailments. Both neurofeedback and biofeedback are starting to be used more frequently by artists, musicians, dancers, and other creative individuals who want to find new ways of connecting people with the world around them, making more immersive experiences. There’s great potential for research in psychology and behavior studies with portable EEG devices that can record brain activity in real-world environments. As the tools for interfacing the brain become more widely available, we will see BCIs come out of medical facilities and labs and become a bigger part of our everyday lives. We envision BCIs revolutionizing everything from neural gaming and augmented reality to meditation and concentration aids. We hope to see OpenBCI lead to toys and tools we haven’t even thought of yet!
Over the past summer we developed working prototypes of the OpenBCI board and verified its functionality and signal quality. We currently have a fully-functional Arduino Shield design, which links the electrical signals of your brain to the limitless power of computation. Now we want to take it to the next level. With your help, we will be able to add wireless bluetooth connectivity to the OpenBCI board to enhance its safety and make it more portable and wearable. We also want to integrate a reprogrammable microcontroller onto the board, as well as a SD card slot for local data storage capability. All of these pieces are tested and in place, we just need your help to leverage manufacturing at scale to make OpenBCI affordable and accessible to everyone. Lastly, like any open-source initiative, the OpenBCI board is only as powerful as the people that use it. We are asking you to join us in revolutionizing the way our minds interface with the world around us by joining the OpenBCI community and letting your voice and mind be heard.
Tools for reading brainwaves have been around since 1912, when Russian physiologist, Vladimir Vladimirovich Pravdich-Neminsky published the first use of electroencephalography (EEG). However, this technology was way ahead of its time. The first digital computers weren’t created until the 1940’s! As a result, the first EEG recordings were difficult to understand, compare, and make relevant. Now, one hundred years later, we live in an age where technology is rapidly getting cheaper, smaller, and more portable. As a result, wearable technology has emerged as an exciting field with seemingly limitless possibilities to record all aspects of life in real-world environments. We think this trend equates to some pretty exciting possibilities for taking EEG outside of the laboratory!
Now, you might be asking yourself what EEG actually means. If you haven’t googled it already, EEG is all about measuring the teeny-tiny electrical signals that our brains make when we do things. These signals are emitted when we think, day-dream, sleep, move around, or meditate - pretty much all the time, if you think about it. Whenever we use our brains, electrical impulses are moving and potentials are flowing all around inside of our heads. EEG is a technique for recording these brain signals.
An EEG system has three basic parts to measure these signals: electrodes which are placed on the scalp; an electronic amplifier that can sense and relay the tiny electrical changes that your brain makes; and a signal processing computer used to make sense of the data and map it to some type of output. After that, the possibilities are endless! One of the most important links in the chain is the amplifier. It is the goal of our Kickstarter to make an open-source, affordable, high-quality EEG amplifier available to everyone so that those possibilities we are talking about can be realized by anyone.
OpenBCI is built around Texas Instrument’s ADS1299 IC. The ADS1299 is an 8-channel, low-noise, 24-bit analog-to-digital converter designed specifically for measuring teeny-tiny EEG signals. It has lots of bells and whistles, like the ability to generate internal signals for testing and calibration, as well as EEG-specific functions like lead off detection, to ensure that the electrodes are making good contact with the subject. It has a programmable Bias signal (DRL) and a very flexible input multiplexer. If you dork out on hardware, like we do, you’ll want to take a look at the datasheet. Because this data sheet can be pretty scary for people that don’t have a background in electrical engineering, we made the OpenBCI board and code libraries so that communicating with and controlling the ADS1299 chip is easy.
The OpenBCI Board comes with an onboard re-programmable micro-controller. All backers that select a reward that includes an OpenBCI Board will get to choose between the 8-bit ATmega328 core made by Atmel (with Arduino bootloader) and a 32-bit PIC core made by Microchip (with chipKIT bootloader). We will send out a survey at the end of the campaign to make sure you receive the microcontroller that works best for your needs. And we have broken out all of the pins so you can blink lights or move motors just like with Arduino or ChipKIT, except now you can do it with your brainwaves! Version 3 of the OpenBCI board will use bluetooth low energy (BTLE) for data transmission and programming of the on board micro-controller. Your safety is super important to us. That’s why we are using a wireless connection to transmit your brain data and the necessary code to and from the OpenBCI board.
But wait, there’s more! The ADS1299 is daisy-chainable. That means you can add another 8 channels to your system by simply plugging on the OpenBCI Daisy Board! Check our rewards list for the OpenBCI Daisy Chain option if you want a lot of electrodes!
The next important part of our OpenBCI system is the electrode sensors. There are two main types: Passive and Active. The OpenBCI board works with both types of electrodes. If you have your own, you can support us at the OpenBCI Board reward level. If you don't have any electrodes lying around, support us at OpenBCI Board w/ Electrode Kit level, we will also send you 10 passive electrodes, conductive paste, and all you need to get started reading your brain waves right out of the box!
In addition to the OpenBCI hardware, we have been working very hard to build template computer programs that visualize and process your brainwaves to make the raw EEG data meaningful. We have code examples built in Arduino, ChipKIT, Processing, Python, and openFramworks. In addition, we have no intention of reinventing the wheel, so we are actively working to make the hardware data accessible to all commonly used open-source EEG signal processing applications, such as BrainBay, OpenVibe, and more. Because you have direct access to the data on the hardware side, making it portable to any existing EEG software is as easy as structuring the way the data is formatted and related. The image below is a screen-shot of our basic OpenBCI Brainwave Visualizer demonstrating 8 channels of raw EEG over a period of 5 seconds and a fast-fourier transform (FFT) plot of the data. This code is available on our Github and can be easily expanded upon.
In the OpenBCI Brainwave Visualizer above, the graphic at the top left shows the electrode placement positions for the experiment. The color intensity of each node indicates the EEG wave amplitude at that region of the scalp. On the right is a graph of the raw EEG data from all 8 channels over time in pretty colors. You can see, at the 4 second mark, a signal artifact in channels 1 & 2 from when our test subject (thanks Aisen!) closed her eyes. The graph in the lower left is an FFT plot, which shows the dominant frequencies in the data stream.
The screenshot was captured about 4 seconds after our test subject closed her eyes. There is a high amplitude waveform seen clearly in channels 5-8, which are placed on the back of the scalp. Notice a spike in the FFT graph at about 10-12Hz. This shows a predominance of alpha waves in the occipital region of the brain, which is what one would expect to see when the subject’s eyes are closed and the visual cortex has nothing to do. Whether you want to do neural therapy to improve your attention or productivity, control video games with your motor cortex, or control the disco lights at your next house party based on your mood, it's the software algorithms and signal processing that make it work. OpenBCI is a community of scientists, engineers, and users coming together to learn and share code and experiences. We hope to make all of these wild ideas become a reality!
In addition to EEG circuit design, we’ve been busy attacking the EEG headwear design challenge from an entirely new angle. Throughout the design process, our goal has been to find a solution that is truly customizable while still taking into account cost, comfort, and signal quality. The reason we think customizability is such a key to the advancement of BCI is because the field is very new and evolving so quickly; there are still so many unknowns in other aspects of the overall BCI design challenge. Low-cost BCI research and development should not be limited by fixed electrode systems that require you to repeatedly sample data from the same regions of the scalp.
At first we considered an injection-molded solution with a semi-fixed form factor but quickly realized this would entail making a lot of assumptions to create a “one-size-fits-all” headset. Inevitably, that design would be too big or too small for some. In addition, because the major cost of injection molding is the design and creation the initial mold, the process would limit iterative design. As a result (and in the spirit of open source) we designed the first 3D-printable EEG headset, with a heavy emphasis on customizability.
Because we want OpenBCI to be beneficial for researchers as well as novice brain hackers, we made sure our design implemented the International 10-20 system – the internationally recognized method for placing electrodes on the human scalp in the context of EEG. We wanted the design to support electrode placement anywhere on the 10-20 diagram, but at the same time, not be bulky and uncomfortable as a result of extraneous components. Because of this, we designed a hierarchical system of snap-in pieces, allowing for a comfortable, personalized headset design.
The body at the back of the 3D-printable headset design (positioned over the visual cortex) has a mount for the OpenBCI Board as well as a slot for a rechargeable lithium battery. The removable arms that extend from the body have nodes where a variety of different hands (1-electrode, 3-electrode, or 5-electrode) can be snapped into place. The option of choosing between 1 and 5 electrodes allows you to target general or granular regions of the scalp, while sticking to 10-20 standards.
To support the assembly and customization of the OpenBCI 3D-printable EEG Headset, we’re developing an online interface that enables you to drag and drop components while visualizing a 3D render of your headset. The interface will also have inputs for important head dimensions (inion-to-nasion distance over the top of the scalp and around the side of the scalp) to algorithmically generate an appropriate headset size. The interface will allow you to export 3D files (in common formats like .obj and .stl) that can be imported into 3D-modeling software like Maya, Inventor, or SolidWorks for further customization. Additionally, you will have the option to order a fully functional pre-printed version of an appropriately sized headset and desired components.
We will be publishing all OpenBCI 3D-Printable Headset files on our Github so that they can be improved by the masses. We hope that the open-source community takes our initial designs and perfects them over time, adapting them for numerous use-cases. 3D printing capabilities are improving rapidly each day and we think the field of BCI has a lot to gain from the potentials of this growing technology.
Research-grade EEG equipment is known to be very cumbersome and very expensive. As hardware has become smaller and cheaper, there are more portable products available that will read your brainwaves and tell you something about your mental state. Today, the leading commercial brain-computer interface companies distribute fixed devices with limited or closed access to the algorithms that translate raw EEG signals into meaningful data. These devices are powerful and effective for application development but not ideal for R&D requiring adjustable hardware setups and direct control over the signal processing techniques. The hardware and software behind OpenBCI are totally transparent; there are no blackbox algorithms or proprietary hardware designs! Not only is it fully accessible, but it is powered by an open-source community of hardware and software builders, making it easy to approach for creators of any skill level and ideal for researchers who modify their system design to suit a specific study. The OpenBCI platform is intended to serve as a malleable tool in the rapidly growing field of brain-computer interfacing.Your donations will allow us to leverage large-scale production and they will support the OpenBCI team to grow and work with the community that you will be helping to found. When you support us at one of the pre-order levels you will receive your OpenBCI hardware in the Spring of 2014, and then our collective efforts to accelerate brain science and research will really kick off! Whether you're a scientist, researcher, hacker, maker, student... your participation will add to the growing pool of available code, implementation tutorials, and user experience. Then we will all be crowd-sourcing brain research! How cool is that!?
The story begins back in the fall of 2011 when Conor was one of Joel’s Physical Computing students in the Parsons MFA Design & Technology program. Conor, who had always enjoyed taking things apart, stumbled upon a hack done by some engineers at NYU’s ITP program. The hack involved taking apart a toy EEG called the Mindflex and porting the brain data into Arduino and Processing. Conor, who was new to neuroscience and electronics, quickly realized the amazing possibilities of interfacing the brain with low-cost, open-source, brain-sensing tools. Joel and Conor worked together to create a hack that allowed EEG data to be stored over extended periods of time by connecting to an Android mobile application through a wearable brainwave-sensing hat.
In completing the project, they both learned a lot about the field of commercial and DIY brain-sensors, but were frustrated by existing systems, due to their limited number of electrodes, fixed headset positions, and the proprietary algorithms that made sense of the raw brain data. Because Joel and Conor were both actively using open-source hardware and software, they longed for a more powerful, open-source alternative.
Then, last winter, out of the blue, Joel got invited to be part of a team applying for a Small Business Innovation Research (SBIR) grant awarded by the Defense Advanced Research Projects Agency to design a portable, high quality, low-cost, open-source, EEG system. The grant was a result of the larger BRAIN Initiative launched by the Obama administration last April. BRAIN stands for Brain Researching Through Advancing Innovative Neurotechnologies, and involves the approximate investment of $100 million "to give scientists the tools they need to get a dynamic picture of the brain and better understand how we think, learn, and remember."
The request to be involved in the project was a crazy coincidence for Joel and he jumped at the chance! Naturally, he got Conor involved due to his passion for the challenges surrounding the field. Now, almost a year later, Joel and Conor have taken their research and development from working under the SBIR grant and turned it into OpenBCI, a fully open-source EEG system for everyone!
Last summer was spent prototyping and testing the hardware circuit designs and building the first versions of the software platforms to visualize the OpenBCI data. The first prototype was unveiled at the World Maker Faire in New York, where Joel and Conor captured the imaginations of folks who had a chance to see their brainwaves for the first time! OpenBCI was honored to receive the Maker Faire Educator’s Choice Award!
Now OpenBCI is hoping to leverage the support of the Kickstarter community to make the OpenBCI board available to the general public. If the Kickstarter campaign is successful, Joel, Conor, and the rest of the growing OpenBCI community will be fulfilling a dream conceived only last year to make an open-source EEG platform available to everyone!
The community that has begun to develop around OpenBCI is one of the main driving forces for us taking the project further. We were very excited by the amazing feedback and support we received at the NYC Maker Faire and it made us realize the importance of getting more people involved.
Moving forward, we want to coordinate many more hackathons and meetups around general techniques of interfacing the brain, showing how OpenBCI and other EEG platforms can be used in tandem to utilize brain data. We love to host hackathons, and if you back us at the Hackathon Reward level, we will coordinate a weekend long event in your town to grow the OpenBCI community and code base. We hope to lower the barrier of entry to the OpenBCI platform so that makers of all ages and skill-sets can contribute to solving the design challenges around putting EEG in real world environments. We plan to build out a well-organized and well-managed forum for sharing questions, ideas, and research. In addition, we hope to see notch hackers emerging who want to share their research and development through an Official OpenBCI Hackers aggregate blog section of our website.
There are already a few early adopters of the OpenBCI technology who are doing really cool things with the system. Our official “Researcher & Developer At Large,” Chip Audette, has been grinding away on some amazing EEG hacks, demonstrating the capabilities of the OpenBCI platform. We hope others will join Chip on leading the way of showing what OpenBCI can really do when there are many minds colliding on the same problems.
We have witnessed the amazing capabilities of open-source platforms like Arduino, ChipKIT, Rasberry Pi, openFrameworks, Processing, and many others, and we are modeling a lot of the OpenBCI structure after their transparent and community-driven practices.
If you help us get funded, we are going to expand our website to enable Chip and others to have a centralized forum for sharing knowledge and R&D techniques. We are in the early stages of developing an online portal for fostering crowd-sourced EEG research and experiments. The site will be a social platform for researchers to meet subjects and share code, experiment protocols, and data. The portal will have 3 major components. First it will serve as a social network for all OpenBCI researchers and developers to connect and share ideas around future research topics and implementations of OpenBCI. Secondly, there will be a platform to create experiments with structured hardware and electrode placement configurations, allowing researchers to crowd-source data collection via a well-structured research initiative. Lastly there will be an organized and completely open repository of data collected by OpenBCI and other EEG hardware.
To help get established and nascent research organizations started with OpenBCI, we are offering an Official Research Partner reward level, which includes an invitation (for up to 5 members of your team) to our Research Partner Conference to be held in the Spring of 2014. Research Partners will also receive 5 OpenBCI Boards or Daisy-Chain Modules, an electrode cap identical to the one seen in our video, and an Electrode Starter Kit. Your entire group will also get beta access to our online portal for collaborative research and data sharing.
Risks and challenges
Product design and development is inherently risky. Bringing a complex technology product to market has many challenges and points of potential breakdown, but we have fully functioning prototypes that we’ve tested in numerous environments. All that’s left is to synthesize the final design, and schedule a production prototype run before mass producing the OpenBCI Board and OpenBCI Daisy-Chain Module.
Most of the design work to bring OpenBCI into production has already been completed. Our current hardware version functions as a shield for Arduino and other Arduino-pin-compatible development platforms. We’ve tested code on the Arduino UNO, DUE, Freescale FRDM board (mbed compiler), and ChipKit Uno32. Software implementations have been designed in Processing, openFrameworks, Python, and MATLAB. We’ve successfully Daisy-Chained two boards and verified up to 16 channels of EEG. The Bluetooth connection has been prototyped and tested. Our final design will be a wireless, re-programmable, 8-channel EEG platform with SD card data-logging and electrode expansion capability.
Joel, one of our artist/engineers, has strong relationships with manufacturers, and we have a good grip on the parts supply-chain, so we hope you will put your trust in us to make good on our promises! We have been working with an established factory in Taiwan that we've used in past projects. They are a large firm with many clients, and they have been helping us throughout every step of the prototyping phase. All of our OpenBCI boards will be pre-programmed and tested before we ship them out to backers.
The great thing about OpenBCI is that it’s totally open source. At this point in time, building on top of the OpenBCI Brainwave Visualizer or building unique applications does require some basic programming knowledge. With that said, our mission is to lower the barrier of entry so that even amateur developers can get up and running right away. With the post-Kickstarter release of the OpenBCI Board V3, we will be supplying thorough starter guides and suggested first project implementations with step-by-step instructions.
As more people adopt the OpenBCI technology, we will encourage our growing community of adopters to give back by sharing their personal research and development methods. We plan to expand the OpenBCI website to include a “Getting Started” section. In addition, we will be coordinating competitions to foster projects that continue to the lower the barrier of entry to interfacing the brain with the OpenBCI platform.
The Electrode Starter Kit does not include the electrode cap seen in the video. In the video we are using Electro-Cap System I sold by Electro-Cap International, Inc. It can be found here: http://www.electro-cap.com/index.cfm/prices-additional-caps-and-supplies/systems/
A similar cap can be found in the following link, sold by Biopac Systems: http://www.biopac.com/electrode-cap-large
These caps are pretty pricy. That is why we are assembling The Electrode Starter Kit, which comes with passive gold electrodes and adhesive conductive electrode paste. This is a great starting place without having to spend a lot of money on the electrode cap. We will be buying the gold electrodes and paste in bulk and assembling cheap starter kits to distribute with the sets.
Every reward that includes and OpenBCI board will also include a battery pack for local power supply.
The OpenBCI board is based on the Texas Instruments ADS1299 chip, which has a sample rate ranging from 250 Hz to 16 Khz. We’ve tested and verified streaming 8 channels of raw, 24-bit EEG over USB serial into a computer at 512 Hz. We’ve tested and verified 16 channels of raw, 24-bit EEG into a computer over USB serial at 256 Hz. As we continue testing and verifying BTLE data rates we will update this FAQ.
Our primary reason for moving away from the Arduino-shield form factor of OpenBCI board V2 is safety. Safety is our number 1 priority, and we want to ensure that we make it physically impossible to ground yourself to a wall while using the OpenBCI platform. To release the Arduino-shield form factor, it would require allocating lots of resources for engineering and testing to ensure safety.
With that said, the OpenBCI board will have exact capability of a standard Arduino and will be programmable through the Arduino IDE. The programming will happen over Bluetooth 4.0 LE instead of USB serial, however.
In short, no. The OpenBCI board will come with a ribbon cable assembly that has the industry standard “touchproof” connectors as shown in the photo in our story. If you are interested in purchasing your own electrodes, there are a number of online distributers that sell different quality EEG electrodes. Not all electrodes will work with EEG systems. Because EEG systems have very low voltage levels (microvolts), you must use electrodes that have very low impedance. Today, most researchers are working with Silver Silver-Chorlide (Ag Ag/Cl) electrodes. Ag Ag/Cl is generally accepted as the best metal to conduct EEG signals.
One of the major design challenges of making EEG data meaningful is looking at how electrical activity in one region of the brain compares to electrical activity in another. The more nodes you have collecting EEG, the better the spatial resolution of your data. By having more electrodes you are able to isolate smaller regions of the brain. With fewer electrodes you are able to look at the brain as whole.
One popular application requiring higher electrode counts is detecting activity in the motor cortex of the brain for classifying the body’s movement. The motor cortex is a thin strip that stretches from ear to ear across the top of the head. When the body is relaxed and in an idle state the motor cortex produces a mu wave (around 13 Hz). This signal is harder to detect than the Alpha rhythms that are easily detectable in the occipital cortex at the back of the head; we demonstrate the ease of detecting alpha in our video and it can be seen in the OpenBCI Brainwave Visualizer in our story.
We foresee a lot of OpenBCI adopters trying to detect and classify the mu rhythms of the motor cortex to build systems that enable virtual control of objects. A system like this will be much more powerful with many electrodes concentrated over the motor cortex, detecting intended movement of one limb versus another. For more information, check out Chip’s comprehensive post on detecting mu rhythms with OpenBCI: http://eeghacker.blogspot.com/search/label/Mu%20Waves
In summary, more electrodes equate to more granular recording of one region of the brain versus another and allows for more complex classification of electrical brain patters.
Our 'Researcher & Developer At Large', Chip, has done some EOG measurements using the OpenBCI V2 board. His experiments can be found here
The OpenEEG Project is an amazing open-source EEG initiative that was started in the early 2000s. The big difference between OpenBCI and OpenEEG is that we are focusing a lot of our efforts on driving down the barrier of entry to brain-computer interfacing, as opposed to just providing an out-of-laboratory option to research-grade EEG. If our Kickstarter campaign is funded, we will be creating numerous how-to guides for novice programmers while still providing a system that is enticing to experienced engineers and programmers. In addition, community outreach is one of our main goals as well. We will be coordinating hackathons geared towards brain-interfacing enthusiasts of all ages and backgrounds. Design, instruction, and approachability are a big part of our vision.
On a more technical note, OpenBCI has a 24-bit ADC instead of OpenEEG being 10-bit. This is a HUGE difference in dynamic range, enabling OpenBCI to do really small signals (EEG), bigger signals (ECG on the wrists), and huge signals (ECG on the chest, EOG, and EMG) all on the same device without clipping or saturation. With OpenEEG, if you do ECG on the wrists, your signals are just barely small enough to fit within OpenEEG's limits. For ECG on the chest, the signals are too big for OpenEEG and your ECG signal shows up on the computer as being clipped. In sum: OpenEEG has a limited dynamic range (10-bits) that is tailored only for EEG. OpenBCI has a wide dynamic range (24-bits) that allows it to support all of these biopotential sensing applications.
We will be hosting a number a few mini-lectures and hackathons over the course of the Kickstarter campaign. On December 16th we will be doing a lecture and live demonstration of the OpenBCI technology at CultureHub in NYC. See event details here: http://www.meetup.com/volumetric/events/153921762/
As more meetups are organized we will be sure to keep all of our backers and the broader OpenBCI community informed.
What's the difference between the 8-bit core Arduino Compatible and 32-bit core ChipKIT Compatible micro-controllers?
If you're wondering which core you should go with, here's a little info to start you off: The 8-bit core runs at 16MHz, and the 32-bit core runs at 50MHz. ATmega328 comes with 32KB of flash program memory space, and the PIC32MX250F128 comes with 128KB of flash program memory.
Specifications on the ATmega328 can be found here
Specifications on the PIC32MX250F128 can be found here
We encourage all backers that are using or plan to use the Daisy-Chain Kit to select the 32-bit PIC32 core. It is a more powerful internal hardware design, capable of faster data rates at higher channel counts. The 8-bit option will be preprogrammed with the Arduino bootloader, making it instantly programmable with all Arduino code. The chipKIT development environment (used with the PIC32 core) is based on the Arduino IDE, but modified to support PIC32. It is compatible with many existing Arduino code examples, reference materials and other resources. Please send us questions that you have through the Kickstarter message portal if you have any. Be sure to do your own research as well, to be certain you maximize your OpenBCI reward!
Support this project
- (42 days)