Telerobotics is the engineering of robotic systems that a human operator controls from a distance, with sensor feedback flowing back so the operator can see, hear, and sometimes feel what the robot encounters. The field spans an enormous range of settings, from a surgeon in New York guiding instruments inside a patient in France, to a NASA engineer nudging a robotic arm on the International Space Station, to a marine biologist piloting a soft gripper along a deep-sea reef. What ties all of these together is a core problem: how do you give a person precise, responsive control over a machine that may be thousands of kilometers away, with all the signal delays, sensory limitations, and safety stakes that distance introduces?
The Latency Problem and How Engineers Work Around It
The single biggest technical headache in telerobotics is latency, the delay between the moment an operator issues a command and the moment the robot executes it. At short range and over fast networks, this delay can be tiny enough that the operator barely notices. But as distance grows, the round-trip signal time becomes impossible to ignore. Light itself takes over a second to travel to the Moon, and signals to Mars face delays measured in minutes. Even on Earth, routing video and control data across continents through fiber optics, satellites, and switching centers introduces measurable lag.
For many industrial or manufacturing applications, researchers have discussed a target of keeping round-trip latency under about one millisecond for smooth, real-time haptic control. When latency creeps above that threshold, operators start to feel a disconnect between their hand movements and the robot’s response, which can cause jerky, unstable motions. In less time-critical tasks, somewhat higher delays are workable, but the general principle holds: the longer the lag, the harder it is for a human to maintain fine motor control through a robotic system.
Space telerobotics pushed engineers to develop creative workarounds. The ROTEX experiment, which flew aboard the Space Shuttle Columbia in 1993, demonstrated a predictive simulation technique for coping with delays of up to seven seconds. The ground station ran a real-time 3D graphics simulation that showed the operator what the robot was expected to do before confirmation came back from orbit. Using this approach, operators on the ground successfully commanded the space robot to grasp a free-floating object automatically, despite the substantial signal lag.1Intelligent Robots and Systems. Telerobotics with large time delays—the ROTEX experience That predictive-display concept remains foundational in space telerobotic design today.
Surgery From Across the Atlantic
The most dramatic public demonstration of telerobotics came in September 2001, when a surgical team in New York performed a cholecystectomy (gallbladder removal) on a patient in Strasbourg, France, using a robotic system connected over a dedicated high-speed fiber-optic line. The operation, often called the Lindbergh Operation, took 54 minutes and went smoothly, with no complications. The round-trip signal covered more than 14,000 kilometers, yet the average transmission delay was only about 155 milliseconds.2PubMed Central. Transcontinental Robot-Assisted Remote Telesurgery: Feasibility and Potential Applications The surgeons reported that the system felt safe and reliable, and the patient recovered normally within two weeks.
That operation was a proof of concept, showing that robot-assisted remote surgery could work across transoceanic distances and that geographical barriers to surgical expertise were not permanent.3Nature. Transatlantic robot-assisted telesurgery More than two decades later, long-distance telesurgery has not become routine, largely because of the infrastructure demands. The Lindbergh Operation relied on a dedicated fiber connection with guaranteed bandwidth and minimal latency. Public internet connections are far less predictable. Still, the underlying technology continues to improve, and shorter-range telerobotic surgery, where the surgeon sits in a console within the same hospital or city as the patient, has become common through systems like the da Vinci platform.
The Missing Sense of Touch
If you have ever tried to pick up a slippery object while wearing thick gloves, you have a rough sense of the challenge telerobotic operators face. Most commercially available surgical robots give the operator excellent visual feedback through high-definition stereoscopic cameras, but they do not transmit meaningful haptic feedback. The surgeon can see the tissue but cannot feel how hard the instrument is pressing against it.4PubMed Central. Methods for haptic feedback in teleoperated robot-assisted surgery
This gap matters. Without force and tactile feedback, surgeons rely entirely on visual cues to judge how much pressure they are applying, which increases the risk of tissue damage. Researchers have developed a variety of sensing and control methods to restore haptic feedback, but practical hurdles remain. Sensors small and biocompatible enough to sit at the tip of a surgical instrument are expensive, and the control loops needed to relay forces back to the operator’s hands without introducing instability are technically demanding. The broader telerobotic field faces a version of this same problem wherever delicate manipulation is involved, whether the operator is handling a coral specimen on a reef or adjusting a component on a satellite.
Exploring the Deep Ocean
Traditional remotely operated vehicles for underwater work use rigid metal grippers that work well for grabbing pipes or cutting cables but are poorly suited for collecting fragile marine life. A rigid claw closing around a sea cucumber or a glass sponge tends to destroy the specimen. Researchers addressed this by developing soft robotic manipulators, essentially flexible grippers made from compliant materials, that an operator can teleoperate with a glove-based interface. In field tests, a soft gripper was successfully operated at depths exceeding 1,800 meters and used to collect organisms including a slime star, a glass sponge, and an extremely delicate holothurian without damaging them.5Scientific Reports. A Dexterous, Glove-Based Teleoperable Low-Power Soft Robotic Arm for Delicate Deep-Sea Biological Exploration
The same system was later deployed from a manned submersible during an expedition to mesophotic coral reefs in Brazil, where it collected branching corals and midwater organisms at depths beyond 700 meters.6Scientific Reports. A Dexterous, Glove-Based Teleoperable Low-Power Soft Robotic Arm for Delicate Deep-Sea Biological Exploration The glove interface is intuitive: the operator simply opens and closes their hand, and the gripper mirrors those movements. This kind of natural mapping between human gesture and robot action is a recurring design goal across telerobotics. The more the control scheme matches what the operator’s body already knows how to do, the less training is required and the fewer errors occur.
Nuclear Disasters and Other Places Humans Cannot Go
Some of the earliest real-world pressure on telerobotic technology came from nuclear accidents. After the Three Mile Island incident in 1979 and the Chernobyl disaster in 1986, robotic systems were sent into radiation-saturated areas where human workers would have received lethal doses in minutes. The Fukushima Daiichi accident in 2011 put those earlier lessons to the test again, and the response revealed both progress and persistent gaps in robotic capability for disaster environments.7Annual Review of Control, Robotics, and Autonomous Systems. The Use of Robots to Respond to Nuclear Accidents: Applying the Lessons of the Past to the Fukushima Daiichi Nuclear Power Station
The challenges in these scenarios are different from surgery or deep-sea biology. The environments are unstructured and unpredictable: rubble, collapsed structures, flooded rooms, extreme radiation that degrades electronic components. Communication links may be unreliable. The robots need to be rugged and somewhat autonomous, because if the link drops momentarily, a robot that freezes in place is better than one that keeps executing its last command. Disaster-response telerobotics has pushed the field toward hybrid systems that blend direct human control with local autonomy, a theme that runs through much of modern telerobotics research.
Shared Autonomy and the Blurring Line Between Control and Collaboration
Pure teleoperation, where every motion of the robot is directly commanded by a human, works when latency is low and the task is well-defined. But in many real situations, the operator benefits from having the robot do some of the work. Shared autonomy is the term for systems where the human and the robot collaborate toward a goal, with the robot inferring what the operator is trying to do and providing assistance.
The tricky part is that the robot often cannot be sure of the operator’s goal until the task is nearly complete. Researchers have developed approaches that let the robot provide useful help even when its confidence about any single goal is low. For example, if the operator might be reaching for one of several objects, the robot can steer the arm in a direction that brings it closer to all of the likely targets at once, narrowing the options as the operator’s intent becomes clearer.8The International Journal of Robotics Research. Shared autonomy via hindsight optimization for teleoperation and teaming This is a significant departure from the simple predict-then-act model, where the robot waits until it is confident about the goal before doing anything helpful.
Shared autonomy is especially valuable for operators with limited mobility or dexterity, such as people using assistive robotic arms, and for situations with communication delays where the robot needs some local intelligence to stay functional between operator commands. The concept also shows up in space telerobotics, where ground controllers send high-level task objectives and the robot handles low-level motion planning on its own.
Telepresence Through Virtual Reality
One of the more futuristic-feeling branches of telerobotics involves immersing the operator in a virtual representation of the remote environment. Mixed-reality telecollaboration systems pair a robotic camera arm at the remote site with a virtual-reality headset worn by the operator. As the operator turns their head, the robotic arm adjusts the camera to match, giving a natural sense of looking around the distant space.9PubMed Central. Viewpoint-Controllable Telepresence: A Robotic-Arm-Based Mixed-Reality Telecollaboration System
A persistent limitation is the restricted field of view and movement range of the robotic camera arm. Researchers have addressed this by combining live stereo video with 3D reconstruction of the surrounding environment, essentially filling in the gaps around the edges of the live feed with a previously captured model of the space. This gives the remote user a wider perception of the local environment and guides them to stay within the range where the camera arm can follow their gaze. Early user studies suggest that VR-based telerobotic interfaces perform at least as well as traditional stereo video displays, and operators tend to prefer them.10Journal of Intelligent & Robotic Systems. The Vicarios Virtual Reality Interface for Remote Robotic Teleoperation
Training People to Operate Robots They Cannot Touch
Learning to operate a telerobotic system is not the same as learning to use a hand tool. The operator has to build an intuition for the robot’s dynamics, its lag, its range of motion, the way the visual feedback maps onto physical space. For surgical telerobotics, this training often happens on simulators before anyone touches a real patient. An interesting finding from training research is that the fidelity of the simulator may matter less than you would expect. A study comparing three different simulator training groups found no significant differences in performance when those trainees moved to the actual da Vinci surgical robot, whether they had trained on high-fidelity or lower-fidelity simulators.11PubMed Central. Simulator Fidelity Does Not Affect Training for Robot-Assisted Minimally Invasive Surgery The result held for both novice and expert operators across suturing and peg-transfer exercises.
This suggests that what transfers from simulator to robot is not a pixel-perfect reproduction of the environment but rather the general coordination skills: learning to control a remote instrument through a console, interpreting stereo video, managing the mental rotation between your hand position and the tool tip’s position in the patient. That is good news for training programs, because lower-fidelity simulators are cheaper and more accessible, especially for institutions in resource-limited settings where telerobotic surgery could eventually have the most impact.
Cybersecurity Risks When the Robot Is on a Network
Any system that operates over a network can, in principle, be attacked over that network. Researchers have explored this specifically for telerobotic surgery, systematically analyzing cybersecurity threats against the Raven II, an advanced teleoperated surgical robot. They demonstrated the ability to maliciously control a wide range of the robot’s functions, and even to completely ignore or override command inputs from the surgeon.12arXiv. To Make a Robot Secure: An Experimental Analysis of Cyber Security Threats Against Teleoperated Surgical Robots
The implications are sobering. An attacker who gains access to the communication channel between surgeon and robot could, in theory, alter the robot’s movements during a live procedure, inject false sensor data, or shut the system down. The Lindbergh Operation used a dedicated private fiber-optic line partly for this reason: the data never touched the public internet. As telerobotic surgery scales to use more widely available networks, cybersecurity becomes a design constraint as fundamental as latency or haptic fidelity. Encryption, authentication, intrusion detection, and fail-safe behaviors (the robot stops if the connection is compromised) are all active areas of engineering work.
Legal Tangles of Operating Across Borders
When a surgeon in one country operates on a patient in another, whose laws apply? This question has no settled answer, and it touches on licensing, malpractice liability, and enforcement of legal judgments. In traditional surgery, jurisdiction follows the location of the procedure. In telesurgery, the surgeon’s hands are in one country and the patient’s body is in another, and it is unclear whether the laws of the surgeon’s country, the patient’s country, or both should govern the procedure.13PubMed Central. The legal and ethical considerations in cross-border telesurgical procedures
Medical licensing is highly localized. A surgeon credentialed in the United States is not automatically authorized to practice in India, even through a robot. Cross-border telesurgery raises the question of whether the surgeon needs a license in the patient’s country, the country where the console is located, or both. Failure to meet licensing requirements could invalidate malpractice insurance, leaving both the surgeon and the patient legally exposed. When something goes wrong, figuring out who is liable becomes even more complicated. If a technical failure causes injury, the fault could lie with the surgeon, the hospital, or the company that built and maintained the robotic system, and different national courts might reach different conclusions about how to distribute that responsibility.14PubMed Central. Liability in Telesurgery: Navigating Legal, Ethical, and Cross-Border Challenges in the Era of Global Surgical Care
Beyond liability, there are ethical considerations around patient autonomy and equity. A patient consenting to telesurgery may not fully understand the technological risks involved, and the informed-consent process for a procedure mediated by a robot, a network, and an operator thousands of miles away is genuinely different from consent for conventional surgery. Meanwhile, the promise that telerobotics could bring specialized surgical care to underserved regions runs up against the reality that the necessary infrastructure, both technical and legal, does not yet exist in most of the places that need it most.
How It All Started
The roots of telerobotics go back further than most people realize. In the mid-1940s, Raymond Goertz at Argonne National Laboratory built the first mechanically coupled master-slave teleoperator, designed to let technicians handle radioactive materials from behind a shielding wall. That first device was purely mechanical: the operator moved a set of handles, and a linked set of tongs on the other side of the wall mirrored the movement. By 1954, Goertz had introduced an electrically driven version with force reflection, meaning the operator could feel resistance when the remote tongs pushed against something. That innovation, giving the operator a sense of the forces the remote manipulator encounters, remains one of the central challenges and aspirations of the field more than seventy years later.
From those origins in nuclear material handling, telerobotics branched into undersea exploration, space operations, surgery, manufacturing, and disaster response. Each domain brought its own constraints and pushed the technology in different directions. Space demanded extreme reliability and tolerance for delay. Surgery demanded precision and minimal latency. Deep-sea work demanded ruggedness and pressure tolerance. The common thread through all of it is the same question Goertz was working on in his lab: how do you extend a person’s hands, eyes, and judgment to a place their body cannot go?