Q2. What are the future works in "Real-time implementation of airborne inertial-slam" ?
Although airborne SLAM is still in its infancy, with many exciting areas for future research, the results presented here have clearly illustrated its capability as a reliable and accurate airborne navigation and mapping system. SLAM consistency and robustness needs to be further investigated.
Q3. What are the key determinants of the feasibility of UAV systems?
Advances in cost effective inertial sensors and accurate navigation aids, such as the Global Navigation Satellite System (GNSS), have been key determinants of the feasibility of UAV systems.
Q4. What is the effect of the SLAM filter on the map uncertainty?
The map uncertainty decreases monotonically towards the lower limit, and it becomes less sensitive to the addition of further information/information gain.
Q5. What was the first flight test of the SLAM system?
After take-off, the vehicle underwent autonomous flight in an oval trajectory, and then SLAM was activated from the ground station.
Q6. What is the relative position vector of the map from the sensor?
The relative position vector of the map from the sensor rssm = [x y z]T is computed from the range, bearing and elevation measurements using a polar-to-Cartesian transformation.
Q7. What are the main issues that need to be resolved?
There are still a number of theoretical, technical, and practical issues that need to be resolved, including SLAM consistency, data synchronisation between vision and INS, real-time implementation of the indirect filter, natural featuredetection and representation, and the incorporation of sub-map techniques for large scale deployment.
Q8. How long does the Brumby airframe take to fly?
The Brumby airframes shown in Fig. 2(a) are capable of flying at approximately 50 m/s and have an endurance of the order of 45 min with a 20 kg payload.
Q9. How can the SLAM algorithm be easily debugged and verified?
By using WindowsTM development tools, the algorithm can be easily debugged and verified, reducing the overall development time needed.
Q10. What is the i th feature position in the navigation frame?
The i th feature position, mni in the navigation frame, is a function of the vehicle position pn , the sensor lever-arm offset from the body centre rbbs , and the relative position of the feature, as measured from the sensor location rssm in the sensor frame [11]:mni = p n +
Q11. What has been done to develop SLAM for 3D environments?
In parallel to these efforts, there have been attempts to develop SLAM for 3D environments, for example: the use of rotating laser range finders in mining applications [8], and the use of stereo vision systems for lowdynamic aerial vehicles [9].
Q12. What are the main reasons why UAVs are so popular?
Unmanned Aerial Vehicles (UAVs) have attracted much attention from robotics researchers in both civilian and defense industries over the past few years.
Q13. What is the DCM for the onboard terrain sensor?
(2)Here Cbs is a DCM which transforms the vector in the sensor frame to the body frame, and is defined for each payload sensor installment.
Q14. How can a SLAM filter be improved?
This can be improved by performing a more precise error analysis, using the errors arising from the inertial sensors and during initialisation.
Q15. How many attempts have been made to implement SLAM in airborne applications?
For airborne applications, to the best of their knowledge there have been only three attempts up to now: SLAM on a blimp-type (thus low-dynamic) platform using a stereo vision system [9]; inertial SLAM in a laboratory environment [10]; and SLAM on a fixed-wing UAV with inertial sensors and a single vision system by the present authors [11].