Backyard Brains Logo

Neuroscience for Everyone!

+1 734 968-7570

items ()

Investigating Cockroach Vision: An Intern Project by U of M Undergrad Greg M.

My name is Greg McMurtry. I am a sophomore at the University of Michigan studying mechanical engineering, and I have been working with Backyard Brains for the past four months. Currently, I am working with cockroaches and their descending contralateral movement detector “DCMD” neuron. Current research suggests that this neuron is correlated with the detection of and reaction to approaching objects.  To test this, we try to take a neural recording from this nerve using a BYB Spikerbox. We then attempt to stimulate the nerve by shining a flashlight at the cockroach to emulate a change in its visual field. My project has two main goals. The first is to develop and refine an experiment that teaches students about the cockroach’s vision and its DCMD. This experiment will be similar in nature to the neural interface surgery used to transform an ordinary cockroach into a RoboRoach. The second purpose of this project is to see how the insects react to certain approaching objects to gain further insight into what the DCMD neuron does and what its response to external stimuli reveals about insect vision.

The procedure used to conduct experiments on the cockroach’s DCMD required that the insect be placed on its back so that electrodes may be inserted through the neck into the DCMD neuron. These electrodes are then connected to the Spikerbox, which can detect any neural activity. In order to gain clear access to the neck, the cockroach’s head was initially anchored back using a piece of string and tape.


This anchoring process was extremely difficult, and it typically took anywhere from five to fifteen minutes to slide the string under the cockroach’s mandible, pulls its head back, and tape the string down without the head shifting during the process. After this process was completed, the cockroach was usually able to move its head into a position that made it very difficult to insert the electrodes. Even if it did not move its head into a difficult position, the cockroach was able to break free from the string and tape within ten minutes of being anchored, leaving a small time frame to conduct the actual experiment.

Frustrated by the difficulty of anchoring the roach’s head back, I began to think of alternative methods to do the same task. Drawing inspiration from the head immobilizers used by emergency services, I began to sketch potential concepts in my lab notebook.


After encouragement from the BYB team, I began designing a more precise model. I began by taking measurements of the cockroach at various points of its body, measurements of the cork, and measurements of the SpikerBox in order to determine the required dimensions for the holder.


When the to-scale model was sketched, I began designing the first model of the Head’s Up! Roach Holder using Sketchup. After printing and assembling the Head’s Up! Roach Holder, I tested it out immediately to see if it worked and if I would need to make changes.


It worked! I was excited that my initial design not only worked, but also far surpassed the string and tape method. It was able to lock both the cockroach’s body and head into place for over an hour, something that the simple tape and string was unable to do.


As a result of this completely immobilized state, I was able to place the electrodes in near perfect position. Initially, a steady beam from a flashlight did not evoke any bioelectrical response. When the flashlight was placed in strobe mode, faint popping sounds were heard coming from the Spikerbox. Upon further testing and combined with visual confirmation from the recordings from the Spikerbox app, these popping sounds were confirmed to be the DCMD neuron processing and reacting to the visual stimulus.


Though significant progress has been made with this project, there is still much more that has to be done. The Head’s Up! Roach Holder will most likely undergo slight revisions so that its height can be adjusted for cockroaches of varying sizes. Future testing will be done in order to qualify the response of the prep based on the electrode positions. The only problem I foresee is determining and verifying the optimal electrode position to get reliable and consistent readings.

Regardless of any possibilities that lie ahead, this research position will continue to be both interesting and rewarding. Check the blog for an update in April!

Gift from Iran thanks to Open Source: cockroach research tools and experiments made by students

On January 1st, we received a New Year’s gift from another continent: Neuroscience tools and experiments made by a group of high school students selected from the 20 best rated schools of Iran. They were written lab reports, submitted for an interdisciplinary neuroscience competition that utilized our open source experiments with cockroaches as a resource for the kids to make their own research and inventions.We here summarize and celebrate their efforts, you can also download the original reports yourself. This is a result of our 3 year friendship with Mohsen Omrani, an Iranian neuroscientist, doing research in nearby Ontario, Canada. He acts as a community liason between the Iran Science communities and the wide network of scientists around the world (Every Iranian Neuroscientist we know seems to be a colleague of Mohsen).


Of note is that in Iran, students choose to follow a biology route or a mathematical root when they are in the 9th grade. There was an emphasis for each team to have students with both biology background and mathematics background so they learn to be able to communicate with each other. So what then did the students investigate?

To begin, a question we often are asked is: “Why Cockroaches?” Indeed this was also asked by members of the Allameh Helli 4 High-school: they submitted the hypothesis that the cockroach is the perfect “explorer” companion for a researcher,  because of their access and survival in complicated and uncertain environments. In other words, they declare that roaches could become better tools than robots for scientists to reach unknown places. The main influences for this conclusion was the article “Line following terrestrial insect biobots” by Tahmid Latif and Alper Bozkurt .

The most remarkable thing about this competition is not that the students built their own tools for the experiments using open source resources, like schematics, code and design… but they made their own custom modifications to design different experiments from the ones we had made.

One excellent example of this is the Robo Roach version (a remote controlled cockroach)  of Alireza Farzad, Behzad Haghgoo, Amir Reza Haji Anzehaei, Aria Hassanpour, Mohammad Reza Osouli of  Allame Helli one High School.

They used an IR System to send a signal to a IR receiver circuit that’s connected to the cockroach antenna AND their cerci. We have only begun cercal stimulation, the Iranian students beat us to it! In words of the students:

“Cerci is a very sensitive organ which receives smallest movements of the air and warns the cockroach to run. We thought that cerci may have a low adaptation rate because it is directly related to its life being. By stimulating the cerci we make an illusion of danger and we make the cockroach run forward”.

Their results to this new experiment was that “ 3V potential difference is the best combination for cerci electric stimulation” and that the cockroach doesn’t adapt to the stimulation of the cerci, unlike the antennas that show strong adaptation properties.

Danial Zohourian and Amir Masoud Azadfar, from a different high school,  focused on cerci stimulation only, coming up with a very useful table of results on how fast the cockroach goes (steps/ per second) according to voltage.

Voltage Recorded Steps Steps per Second
0.5 10 steps in 7 seconds 1.42
1.0 9 steps in 3 seconds 3.00
1.5 12 steps in 4 seconds 3.00
2.0 13 steps in 3 seconds 4.33
2.5 10 steps in 5 seconds 2.00
3.0 13 steps in 4 seconds 3.25
3.5 No Respond Adapted

Interestingly, they had a different outcome than the students from Allame Helli one High School: they concluded that  best stimulation is at 2 volts, not 3, and that cercal stimulation does adapt.

So what is the correct answer?  Only that new experiments are necessary to understand why there are different results, and what improvements are important to obtain a more accurate conclusion. But as we have learned, the best experiments come from disputes between scientists that motivate each other to improve their work.

Regarding on this emphasis on possible errors to improve experiments, the writing of  students Tarannom Taghavi and Nastaran Fatemi, from Kherad high school caught our attention. They tried  to tackle the main problem of the Roboroach: the behvioral adaptation to the stimuli that controls the cockroach:  “ If we can produce the signals in it’s ganglion and send it to the cockroach, there won’t be adaptation anymore.  As we are creating the signals and sending it to its decision making  center, we might be able to take control of cockroach’s decision  making process.” They did this by recording roaches signal with a spikerbox and trying to send it back to the ganglia.


Interpretation of the electric signal obtained from the cockroach.

 Although it wasn’t successful, coming up with this hypothesis to solve the main problem of RoboRoaches was impressively creative. And, as we noted, we really liked the focus of their paper in the mistakes that were made and how to make corrections for a future experiment: they were the only students that made emphasis on the importance of iteration, of making a lot of failed experiments that are patiently and constantly improved, before making any discovery. Thus our informal “Golden Cockroach” award goes to Tarannom Taghavi and Nastaran Fatemi.

Finally, we want to give a special mention to the only group that designed a new interface: a special cockroach treadmill  to estimate the adherence of these insects legs:


Keep on inventing, Keep on discovering, our fellow young colleagues across the globe.

You can download the original writings here and see the competition video below

The video cannot be shown at the moment. Please try again later.

Backyard Brains visits China

This past August, we were graciously invited by ZAST (Zhejiang Association for Science and Technology) to come to China to give a series of workshops and talks in Hangzhou, Ningbo, and Shanghai at the various impressive science museums there. Similarly to the United States, Neural Engineering is a relatively new discipline for the public, and there was enormous interest in the topic.

We were honored. In early August, Tim flew to Hangzhou and began the Sino-Backyard Brains adventures. At the beginning of the trip, we gave a “deep workshop” where the staff of the Zhejiang Museum of Science and Technology (ZMST), affiliated to ZAST, received 6 hours of training on interactive experiments, ranging from our cockroach work, to our earthworm conduction velocity experiments, to our new human interfaces, and even our (then just prototyped) EEG experiments.

Hangzhou is famous for its West Lake, and it was quite beautiful walking around the city and visiting the nearby Dragonwell (Longjing) tea farms, all the while planning the subsequent science talks with Bing, who was the organizer of the visit.

One of the highlights was to give an ASTalk at the ZMST. The audience consisted of a mix of students from local high schools, universities, and the public. They were treated to the first demos of our robot hand interface, a Chilean-USA collaboration between Backyard Brains and the Chilean Startup “HackerHand.”


This was followed by a organized field trip to the Zhuxiang National Park with grammar school students, the ZAST’s museum staff, and an entomologist. The objective was to try the RoboRoach preparation on a Chinese Beetle (unidentified, perhaps a dung beetle). Zhuxiang is known for being one of the best examples of a bamboo forest.

We searched for our elusive beetle in the Chinese scenery borne of dreams.

The beetles were found on a distinctive tree that was co-localized with a type of plentiful brown butterfly easy to spot. Both are fond ofthe sap the tree emits. You look for the butterfly, then look for the tree the Butterfly always flies to, and you find your beetle.

The video cannot be shown at the moment. Please try again later.

We collected about 10 of these beetles, and implanted 3 of them with RoboRoach electrodes. These beetles have an impressively strong exoskeleton, making the surgery a bit of a challenge, but we achieved it.

However, the wires did not hold up to the Beetles’ jaws. The next morning, when we tried to test the RoboRoach circuit, the beetle had cut the wires. The surgical preparation needs a little bit more adaptation to work on this creature, but we left the RoboRoach kit with ZAST…maybe an adventurous Chinese student will continue the work to study the adaptation and motor response properties of this beetle!

We gave many talks to the public, who were very gracious, and we must thank Phyllis, Bing, and our talk translators for doing the challenging job of learning so many new biological terms. Our favorite was- “ulnar nerve” – chî gû shén jin.

The translators taught us two words for the cockroach. There is the formal name “cockroach” which sounds like zháng láng

But there is also an informal word used that translates into “little strong” because, of course, a cockroach is a quite durable creature– which sounds like xiâo qiáng.

The world is full of enthusiastic students and minds who live to unravel the mechanisms of the brain, and we were delighted to meet some of the curious future Neural Engineers in China. The Staff of ZAST, when we parted ways in Shanghia, gave a gift all engineers would treasure – an abacus. The Abacus is now part of our office in the Santiago MakerSpace in Chile, in our growing ganglious network of inventors and scientists spread around the world.