![Current RF localization schemes are too coarse
• State-of-the-art WiFi localization: 23cm
[ArrayTrack]
• State-of-the-art RFID localization: 11cm [PinIt]
BUT requires a dense grid of reference tags
How to get a few cm accuracy without
environment instrumentation?](https://image.slidesharecdn.com/rfcompass-150312100030-conversion-gate01/85/Rf-compass-10-320.jpg)
![• Compare distances between RFIDs
Which blue tag is closer to the red tag?
Tag 3
Tag 1
Tag 2
Distance ordering based on signal similarity
[SIGCOMM’13]
Building block: RF pairwise comparison](https://image.slidesharecdn.com/rfcompass-150312100030-conversion-gate01/85/Rf-compass-14-320.jpg)
Rf compass
- 1. RF-Compass: Robot Object Manipulation Using RFIDs Fadel Adib, Ross Knepper, Dina Katabi, Daniela Rus Jue Wang
- 2. Limitation of Today’s Robotic Automation Fixed-position, single-task robot • Limited to large-volume production line • Inability to change manufacturing process
- 3. Toyota has been slowly backing away from heavy automation. The labor saved by robots was wasted most of all by reprogramming robots. This is the future. A new wave of robots, far more adept than those now commonly used by automakers and other heavy manufacturers. The potential for much broader industrial acceptance is tied to the development of robots that can absorb data, recognize objects, and respond to information and objects in their environment with greater accuracy.
- 4. Mobile Manipulation Fetching, grasping, and manipulating objects • Extend automation to small/medium factories • Easy to reconfigure manufacturing process
- 5. • Centimeter-scale localization, e.g., 2cm • Minimal instrumentation portable Requirements for Mobile Manipulation
- 6. Current Approaches • Motion capture system, e.g., VICON – Sub-centimeter accuracy – Heavy instrumentation & Expensive
- 7. Current Approaches • Motion capture system, e.g., VICON – Sub-centimeter accuracy – Heavy instrumentation & Expensive • Imaging (e.g., optical camera, Kinect, LIDAR) – Needs prior training or ?
- 8. Current Approaches • Motion capture system, e.g., VICON – Sub-centimeter accuracy – Heavy instrumentation & Expensive • Imaging (e.g., optical camera, Kinect, LIDAR) – Needs prior training or ?
- 9. Can RF localization help?
- 10. Current RF localization schemes are too coarse • State-of-the-art WiFi localization: 23cm [ArrayTrack] • State-of-the-art RFID localization: 11cm [PinIt] BUT requires a dense grid of reference tags How to get a few cm accuracy without environment instrumentation?
- 11. RF-Compass • Place RFID tags on both robot and objects • No reference tags in the environment
- 12. Identifying the Object • RFID: a passive sticker – no battery, low cost • Reader shines RF signal on tags Each tag replies with its unique ID Works for up to 10 meters
- 13. How to get centimeter-scale accuracy?
- 14. • Compare distances between RFIDs Which blue tag is closer to the red tag? Tag 3 Tag 1 Tag 2 Distance ordering based on signal similarity [SIGCOMM’13] Building block: RF pairwise comparison
- 15. Basic building block 2cm accuracy
- 16. Basic Idea: Localization by Partitioning Is the red tag closer to Tag 1 or Tag 2?
- 17. Basic Idea: Localization by Partitioning Tag 1 is closer than Tag 2
- 18. Basic Idea: Localization by Partitioning Tag 3 is closer than Tag 4
- 19. Basic Idea: Localization by Partitioning Tag 4 is closer than Tag 1
- 20. Basic Idea: Localization by Partitioning But not yet centimeter accuracy
- 21. Basic Idea: Localization by Partitioning • Partitions can be iteratively refined
- 22. Iterative Refining via Robot Navigation • Leveraging robot’s consecutive moves
- 23. Iterative Refining via Robot Navigation • Every robot move gives a new set of partitions
- 24. Iterative Refining via Robot Navigation • Lay new partitions over old partitions to refine
- 25. • Keep refining until reaching centimeter accuracy Iterative Refining via Robot Navigation
- 26. • Keep refining until reaching centimeter accuracy Iterative Refining via Robot Navigation
- 27. Formulation as an Optimization 2 𝑥2 − 𝑥1 2 𝑦2 − 𝑦1 𝑥0 𝑦0 ≤ 𝑥2 2 + 𝑦2 2 − 𝑥1 2 − 𝑦1 2 (𝑥1, 𝑦1) (𝑥2, 𝑦2) (𝑥0, 𝑦0)
- 28. Formulation as an Optimization 2(𝑥2 − 𝑥1) 2(𝑦2 − 𝑦1) ⋮ ⋮ 𝑥0 𝑦0 ≤ 𝑥2 2 + 𝑦2 2 − 𝑥1 2 − 𝑦1 2 ⋮ (𝑥0, 𝑦0)
- 29. Formulation as an Optimization 𝑨 𝑥0 𝑦0 ≤ 𝒃 Works correctly even if randomly flipping 10% of pairwise comparisons, shown in paper (𝑥0, 𝑦0) • A feasibility problem with linear constraints • Efficiently solved via convex optimization • Over-constrained system ↓ Robustness to errors & outliers
- 30. Problem: also need orientation for grasping Solution: • Multiple RFIDs on object • Naïve approach: localize each RFID independently and find orientation • Our approach: joint optimization using knowledge of their relative location Orientation
- 31. Evaluation • Used a robot to fetch IKEA furniture parts • 9 tags on robot, 1 – 4 tags on object
- 32. Baseline • VICON motion capture system • Sub-centimeter accuracy • Infrared cameras + infrared-reflective markers VICON Markers
- 33. Navigation Performance CDF CDF Ratio to Optimal Path in LOS Ratio to Optimal Path in NLOS Direct line-of-sight RF-Compass enables effective navigation in NLOS VICON does NOT work in NLOS Occlusion and NLOS Only 6% longer than optimal on average
- 34. Center Position Accuracy 4 cm 2.8 cm 1.9 cm 1.3 cm 0 1 2 3 4 5 6 1 Tag 2 Tags 3 Tags 4 Tags Number of Tags on Furniture Part ErrorinPositionEstimate(cm)
- 35. Number of Tags on Furniture Part ErrorinOrientation(degree) 5.8˚ 3.6˚ 3.3˚ 0 1 2 3 4 5 6 7 2 Tags 3 Tags 4 Tags Orientation Accuracy
- 36. Conclusion • RF-Compass: accuracy of a few cm and degrees • Iterative refining by leveraging robot’s navigation • Opens up opportunities for bridging robot object manipulation with RF localization