A telepresence robot is a remotely operated mobile robot that carries a display, cameras, microphones and speakers, allowing a person in another location to see, hear, speak and move through a physical space as if present. In robotics, it sits between videoconferencing and autonomous service robots: the human supplies the judgment, while the machine supplies mobility and a physical presence.
Unlike a fixed video call, a telepresence robot lets the remote user choose where to look and where to go, join conversations in a corridor or inspect equipment on a shop floor. Most units combine a wheeled base with shared autonomy, meaning the operator issues high-level commands while onboard software avoids obstacles and handles docking.
How a telepresence robot works
A telepresence robot works as a closed loop between a remote operator and a mobile platform: video and audio stream from the robot to the operator, and driving commands stream back over the network. The operator usually uses a web browser, desktop application or tablet, and the robot connects through Wi-Fi or a cellular link to a cloud relay or a direct peer-to-peer session.
Real-time media normally relies on protocols such as WebRTC, which adapt bitrate to available bandwidth. The onboard robot controller translates velocity commands into motor outputs, reads safety sensors and enforces speed limits, so that a dropped connection stops the robot rather than leaving it moving.
The table lists the main subsystems and what each contributes to the experience of being present.
| Subsystem | Typical components | Role in telepresence |
|---|---|---|
| Mobility base | Differential drive, two to four wheels | Moves the operator through the site |
| Head unit | Display, wide-angle camera, speaker | Shows the operator and captures the view |
| Audio array | Several microphones with echo cancellation | Picks up voices in noisy rooms |
| Navigation sensors | Lidar, depth cameras, ultrasonic, cliff sensors | Obstacle and drop-off avoidance |
| Connectivity | Wi-Fi, LTE or 5G modem | Carries video and control traffic |
| Docking station | Charging contacts, alignment markers | Recharges between sessions |
Key parameters and design trade-offs
The usefulness of a telepresence robot depends on a handful of measurable parameters, and improving one usually costs another. End-to-end latency matters most: below roughly 200 milliseconds driving feels direct, while delays above 500 milliseconds make operators overshoot turns and collide with obstacles.
Typical values below reflect commercial and research units; they vary with network quality and design priorities.
| Parameter | Typical range | What it affects |
|---|---|---|
| Glass-to-glass latency | 150 to 600 ms | Driving precision and conversation flow |
| Uplink bandwidth | 1 to 4 Mbit/s for HD video | Image clarity and frame rate |
| Top speed | 0.5 to 1.5 m/s | Ability to keep pace while walking |
| Battery runtime | 2 to 12 hours | Length of continuous sessions |
| Display height | 1.2 to 1.7 m, often adjustable | Natural eye level in conversation |
| Robot mass | 5 to 50 kg | Stability, safety and portability |
A taller head improves eye-level contact, which strongly influences how colleagues treat the remote person, but raises the center of gravity and the risk of tipping. Lightweight designs are cheaper and safer around people, yet they struggle with thresholds, thick carpets and ramps.
Telepresence robots versus related robot types
A telepresence robot is a specialized service robot whose primary payload is human communication rather than physical work. Its closest relatives differ mainly in who makes the decisions and whether the robot manipulates objects.
| Robot type | Who decides | Main purpose | Manipulation |
|---|---|---|---|
| Telepresence robot | Remote human, with assistance | Communication and remote presence | Rarely |
| Autonomous mobile robot | Onboard planning software | Transport and logistics | Sometimes |
| Teleoperated manipulator | Remote human, continuously | Handling objects at a distance | Always |
| Humanoid | Mostly autonomous, sometimes teleoperated | General tasks in human spaces | Usually |
| Drone | Pilot or autopilot | Aerial observation and delivery | Rarely |
The line is blurring. Some autonomous mobile robots offer a telepresence mode for remote inspection, and some humanoid platforms are teleoperated to collect training data. The defining feature remains that a human is in the loop for the social interaction.
Where telepresence robots are used and how to choose one
Telepresence robots are most valuable where a remote person needs to move between locations inside a building and interact spontaneously, rather than attend a single scheduled meeting. The decision table matches common settings with suitable features.
| Setting | Recommended features | Reason |
|---|---|---|
| Hospital rounds | High-resolution zoom camera, auto-docking | Clinicians read monitors and charts remotely |
| School attendance | Simple controls, adjustable height | Students with illness join classes and peers |
| Office collaboration | Good microphones, long battery life | Remote staff join corridor discussions |
| Factory inspection | Robust base, cellular backup, lights | Experts assess equipment without travel |
| Museum visits | Guided routes, accessibility options | Distant visitors explore exhibits |
Before deployment, teams should verify Wi-Fi coverage along every route, since roaming gaps between access points cause frozen video and emergency stops. Elevators, heavy doors and steps usually require assisted access or building automation integration.
Limitations, risks and common misconceptions
The most common misconception is that a telepresence robot replaces physical presence entirely. It conveys sight and sound well, but it cannot shake hands, open most doors or handle objects, and conversational cues such as gaze direction are only partly preserved.
- Network dependence: quality collapses on congested or patchy wireless networks.
- Privacy concerns: a moving camera in hospitals, schools or homes requires clear consent rules and visible session indicators.
- Security exposure: remote access accounts must use strong authentication, because a hijacked unit becomes a mobile surveillance device.
- Social acceptance: people may ignore or talk over the robot unless the remote user has authority and eye-level presence.
- Physical safety: low-mass units are generally safe, but collisions on stairs or with children demand cliff sensors and speed limits.
A plain-language summary
A telepresence robot is like a video call on wheels. A screen showing a person’s face sits on top of a small rolling base. The person, who may be in another city, steers it from a laptop and sees through its camera.
This lets someone visit a classroom, walk through a hospital ward or join a chat by the coffee machine without traveling. The robot handles the moving around and avoids bumping into things, while the human does the talking and deciding.
An example from practice
A regional hospital network has one stroke neurologist on night duty for five smaller hospitals. Each emergency department keeps a telepresence robot docked near the resuscitation bay, connected over the hospital Wi-Fi with a cellular backup link.
When a patient arrives with stroke symptoms, a nurse wakes the robot and the neurologist logs in from home within minutes. The specialist drives the robot to the bedside, uses the optical zoom camera to observe facial symmetry and pupil response, speaks with the patient and family, and reviews scans on a second screen.
The neurologist then advises on clot-dissolving treatment while the local team acts. The network reports that specialist assessment starts far sooner than when patients were transferred first, which matters because stroke outcomes depend heavily on time to treatment.