"Long before it's in the papers"
January 28, 2015


Monkey controls robotic arm with thoughts

May 28, 2008
Courtesy U. of Pittsburgh
and World Science staff

Us­ing just their thoughts, mon­keys have car­ried out smooth, well-con­trolled move­ments of a robotic arm to feed them­selves, re­search­ers say.

The de­vel­op­ment is the lat­est ap­plica­t­ion of a new tech­nol­o­gy—brain-com­put­er in­ter­faces, or BCI—meant to ben­e­fit peo­ple with spi­nal cord in­ju­ries or pa­ral­y­sis. In these sys­tems, com­put­ers in­ter­pret some­one’s in­tent­ions based on elec­tri­cal sig­nals cours­ing through the brain. The ma­chines then use the in­formation to con­trol elec­tron­ic de­vices.

A monkey feeds it­self a marsh­mal­low us­ing a ro­bot­ic arm. A vi­deo is here (Win­dows Me­dia Play­er file).

Pre­vi­ous BCIs were used for sim­pler tasks, such as con­trolling cur­sor move­ments on a com­put­er screen. The newer ver­sion with the “arm” is meant to be more ap­plicable to real-life situa­t­ions.

“Our im­me­di­ate goal is to make a pros­thetic de­vice for peo­ple with to­tal pa­ral­y­sis,” said neu­ro­bi­olo­g­ist An­drew Schwartz, sen­ior au­thor of a pa­per in the May 29 is­sue of the re­search jour­nal Na­ture de­tail­ing the find­ings. 

“Ul­ti­mately, our goal is to bet­ter un­der­stand brain com­plex­ity,” added Schwartz, of the
Uni­ver­s­ity of Pitts­burgh School of Med­i­cine.

Mon­keys in his lab moved a robotic arm to feed them­selves marsh­mal­lows and chunks of fruit while their own arms were re­strained. In the de­vices, spe­cial soft­ware reads sig­nals pick­ed up by elec­tri­cal probes the width of a hair. These are in­sert­ed in­to sig­nal­ing path­ways of neu­rons, or brain cells, in a brain re­gion where vol­un­tary move­ment orig­i­nates as elec­tri­cal im­pulses.

The neu­rons’ col­lec­tive ac­ti­vity is then eval­u­at­ed us­ing soft­ware pro­grammed with spe­cif­ic for­mu­las and then sent to the arm, which exe­cutes the ac­tions the mon­key meant to per­form with its own limb. The move­ments are flu­id were nat­u­ral, and ev­i­dence shows the mon­keys come to re­gard the robotic de­vice as part of their own bod­ies, re­search­ers said. A vi­deo of a mon­key oper­at­ing the arm is here.

The pri­ma­ry mo­tor cor­tex, a brain re­gion that con­trols move­ment, has thou­sands of neu­rons that sig­nal to­geth­er as they con­trib­ute to cre­at­ing mo­tion. Be­cause so many neu­rons fire to­geth­er for even the sim­plest ac­tion, it would be im­pos­si­ble to cre­ate probes that read each cel­l’s sig­nals. Schwartz’s group de­vel­oped a for­mu­la that uses lim­it­ed in­forma­t­ion from only 100 neu­rons to fill in the mis­sing sig­nals.

“The mon­key learns by first ob­serv­ing the move­ment, which ac­tivates his brain cells as if he were do­ing it,” he said. “It’s a lot like sports train­ing, where train­ers have ath­letes first im­ag­ine that they are per­forming the move­ments they de­sire.”

In a pre­vi­ously de­vel­oped brain-com­put­er in­ter­face in which peo­ple con­trolled cur­sors, no wires at all were placed the brain. The de­vice was de­signed to simply read elec­tri­cal ac­ti­vity around the brain gen­er­at­ed by its in­ter­nal cel­lu­lar sig­nal­ing. The ad­van­tage of the newer de­vice is that while it still re­lies on the brain probes, it al­lows three-di­men­sion­al move­ment, said Schwartz: “We’ve dem­on­strat­ed a high­er lev­el of pre­ci­sion, skill and learn­ing.”

* * *

Send us a comment on this story, or send it to a friend


Sign up for

On Home Page         


  • St­ar found to have lit­tle plan­ets over twice as old as our own

  • “Kind­ness curricu­lum” may bo­ost suc­cess in pre­schoolers


  • Smart­er mice with a “hum­anized” gene?

  • Was black­mail essen­tial for marr­iage to evolve?

  • Plu­to has even cold­er “twin” of sim­ilar size, studies find

  • Could simple an­ger have taught people to coop­erate?


  • F­rog said to de­scribe its home through song

  • Even r­ats will lend a help­ing paw: study

  • D­rug may undo aging-assoc­iated brain changes in ani­mals

Using just its thoughts, a monkey has carried out smooth, well-controlled movements of a robotic arm to feed itself, researchers say. The development is the latest application of a relatively new technology, brain-computer interfaces or BCI, in which machines read electrical signals coursing through the brain to control electronic devices. Previous BCIs were used for simpler tasks, such as controlling cursor movements on a computer screen. The newer version with the “arm” is meant to be more applicable to real-life situations. This could benefit people with spinal cord injuries or paralysis, according to the researchers, at University of Pittsburgh School of Medicine. “Our immediate goal is to make a prosthetic device for people with total paralysis,” said Pitt neurobiologist Andrew Schwartz, senior author of a paper in the May 29 issue of the research journal Nature detailing the findings. “Ultimately, our goal is to better understand brain complexity.” Monkeys in his lab moved a robotic arm to feed themselves marshmallows and chunks of fruit while their own arms were restrained. In the devices, special software interprets signals picked up by electrical probes the width of a hair. These are inserted into signaling pathways of neurons, or brain cells, in a brain region where voluntary movement originates as electrical impulses. The neurons’ collective activity is then evaluated using software programmed with specific formulas and then sent to the arm, which carries out the actions the monkey meant to perform with its own limb. The movements are fluid and natural, and evidence shows the monkeys come to regard the robotic device as part of their own bodies, researchers said. The primary motor cortex, a brain region that controls movement, has thousands of neurons that signal together as they contribute to creating motion. Because so many neurons fire together for even the simplest action, it would be impossible to create probes that read each cell’s signals. The Pitt researchers developed a formula that uses limited information from only 100 neurons to fill in the missing signals. In a previously developed brain-computer interface system in which people controlled cursors, no wires at all were placed the brain. The device was designed to simply read electrical activity around the brain generated by its internal cellular signaling. The advantage of the newer device is that while it still relies on the brain probes, it allows three-dimensional movement. “We’ve demonstrated a higher level of precision, skill and learning,” said Schwartz. “The monkey learns by first observing the movement, which activates his brain cells as if he were doing it. It’s a lot like sports training, where trainers have athletes first imagine that they are performing the movements they desire.”