In recent years, the market for service robots and intelligent mobile devices has been evolving rapidly. Follow-me technology has moved beyond being a mere concept and is now powering real-world products. From autonomous luggage in airports to smart shopping carts in retail stores and assistive wheelchairs for seniors, the core capability behind all these innovations is the same: how to reliably identify and track a user while navigating in dynamic environments.
Currently, the mainstream positioning and sensing technologies for follow-me applications can be grouped into three categories:
- UWB (Ultra-Wideband positioning)
- Vision (camera-based perception and AI recognition)
- LiDAR (Light Detection and Ranging)
But how do these technologies compare? What are their strengths and limitations? And, most importantly, which one is best suited for specific applications?
UWB: A High-Precision “Invisible Tether”
Advantages
- High accuracy: Can achieve positioning accuracy of 10–20 cm.
- Robustness to environment: Works reliably in dim lighting, crowded areas, or when direct line of sight is blocked.
- Low latency: Ensures real-time response, which is critical for strollers, wheelchairs, or safety-related applications.
Limitations
- Requires the user to carry a tag or beacon, which may slightly affect user experience.
- Indoor environments with strong reflections or obstructions may require additional anchors or calibration.
Typical Applications
- Follow-me baby strollers and wheelchairs
- Material transport carts in factories
- Golf bag trolleys
In one sentence: UWB acts like an invisible rope, keeping the vehicle tightly connected to the user.
Vision: AI-Powered “Smart Eyes”
Advantages
- No additional device needed: Users don’t need to wear a tag; the system recognizes them directly through cameras.
- Rich recognition capability: Can identify not only the user but also obstacles, landmarks, or even gestures.
- Decreasing cost: Lightweight AI models (e.g., YOLO, MobileNet) now run on compact, low-power hardware.
Limitations
- Sensitive to environment: Prone to failure in poor lighting, strong backlight, or crowded situations.
- Higher compute demand: Requires AI inference hardware for real-time recognition.
- Limited field of view: Occlusion or sharp turns may cause tracking loss.
Typical Applications
- Smart shopping carts in retail stores
- Autonomous luggage in airports and train stations
- Entertainment or camera-following robots
In one sentence: Vision offers the most natural way of following, but it can be easily disrupted by real-world conditions.
LiDAR: The Precision “Environment Scanner”
Advantages
- Excellent environmental perception: Can generate precise 2D/3D maps and detect obstacles with high accuracy.
- Stable performance: Works consistently regardless of lighting or time of day.
- Supports SLAM: Enables follow-me combined with autonomous navigation.
Limitations
- High cost: Advanced LiDAR sensors are still relatively expensive.
- High computational complexity: Requires powerful processors for real-time SLAM and decision-making.
- Weak at user identification: LiDAR excels at mapping environments, but not at recognizing individual people.
Typical Applications
- Material transport in industrial parks
- Automated Guided Vehicles (AGVs) in warehouses
- Security patrol robots
In one sentence: LiDAR is the expert in environmental awareness, but needs to be paired with other technologies for human-centered following.
Which Technology Is the Best Choice?
There is no single “best” technology. The optimal choice depends entirely on the application:
1. Human-centered follow-me (baby strollers, wheelchairs, luggage)
UWB + Vision fusion: UWB ensures stable relative positioning, while vision provides user recognition and obstacle avoidance.
2. Retail and commercial scenarios (shopping carts, guide robots)
Vision-based with LiDAR support: Vision recognizes the customer, and LiDAR ensures safe navigation and collision avoidance.
3. Industrial and logistics (factory transport carts, warehouse robots)
LiDAR-based with UWB support: LiDAR handles mapping and path planning, while UWB helps locate users or specific workstations.
The Future: Fusion Is the Answer
As AI computing becomes more efficient and sensor costs continue to fall, relying on a single technology is no longer sufficient. The future of follow-me technology lies in multi-sensor fusion:
- UWB provides precise relative positioning.
- Vision enables user recognition and environmental understanding.
- LiDAR guarantees safe navigation and robust obstacle avoidance.
This hybrid approach will allow follow-me vehicles to work reliably across diverse environments – from homes to factories, indoors to outdoors.
Conclusion
At its core, follow-me technology is about helping machines better understand where people are and how to assist them.
- UWB is the precise “positioning assistant.”
- Vision is the intelligent “pair of eyes.”
- LiDAR is the reliable “environment scanner.”
The best solution is not to choose one over the others, but to combine them strategically based on the scenario. This convergence will be the key differentiator – and the driving force behind the next generation of follow-me products.