3 ms·

That's some pretty pathetic collision avoidance, given that the website advertises "Its built-in guidance system takes care of the rest by avoiding furniture &

That's some pretty pathetic collision avoidance, given that the website advertises "Its built-in guidance system takes care of the rest by avoiding furniture & people..."

palish 16y ago
Really? It experienced the worst-case scenario -- running directly into something at full speed -- and came away completely unharmed. "Rock solid" would be how I'd describe it. It would be very tough to implement "collision avoidance" without being overbearing on what the user can or cannot do. For example, it would be extremely annoying if the robot's collision avoidance made it actively try to stay away from walls "because the user could potentially make a sudden turn at full speed towards the wall, and then a collision could result, so therefore we should try to avoid that scenario because, damn it, we promised collision avoidance!". The net effect would be that the user might have a very good reason for wanting to drive close to a wall, but can't because of artificial limits imposed by the software. It's better to not waste time on those kinds of 'features' and instead focus on making the core experience better. In this case, focusing on the two-way video and reducing chat lag time. Incidentally, those kinds of 'features' seem like exactly the kind of thing that a mature company would focus on (waste time on) and lose sight of making the actual product better, because it combines so many things that mature companies would like -- delivering on promises to the user, it sounds good on paper, etc. But it loses sight of the fact that the goal is to make the end product better, not to deliver a laundry list of things.
- pavs 16y ago
  Why does the collision avoidance has to be a hard limit? Allow it to be toggled on and off at will by the operator and it solves all problems you mentioned. Collision off on cramped places with furniture. Collision on when too many moving object (people walking around).
  - robobenjie 16y ago
    Actually we allow people to hit walls, for the reasons you mention above, but we make sure that the robot slows down a little before hitting them. Basically we allow any movement but you can only go fast when you are in an open area and if the robot detects that you are going to barely clip an obstacle you barely miss it instead (thats the hope anyway) -Benjie, anybots obstacle avoidance guy
    - spoon16 16y ago
      what happens if I drive it down stairs?
      - robobenjie 16y ago
        It depends on how long the stairway is. It has a bumper around the head to protect it, but odds are you are going to need some repairs. We are using the Hokuyo Lidar sensor: http://www.hokuyo-aut.jp/02sensor/07scanner/urg_04lx_ug01.html http://www.hokuyo-aut.jp/02sensor/07scanner/urg_04lx_ug01.ht... which sees an ankle height slice of the world so stairs down are invisible to us at the moment. :/
        tudorachim 16y ago
        Hey, I think my original post was totally misinterpreted. I was referring to the instance near the beginning where Arrington drove the robot forward, directly into Dan Casner, a few times. That mistake will almost certainly be repeated by everybody who uses QA, as they try to get a better view of the people they're talking to. So, I was commenting on the lack of a hard limit on how close one can get to objects directly in front of the robot. Will this be addressed in the final implementation? I believe there might also be a ROS package that detects ankles (well, legs) in 2D scans, which could help.

That's some pretty pathetic collision avoidance, given that the website advertises "Its built-in guidance system takes care of the rest by avoiding furniture & people..."

palish 16y ago

Really? It experienced the worst-case scenario -- running directly into something at full speed -- and came away completely unharmed. "Rock solid" would be how I'd describe it. It would be very tough to implement "collision avoidance" without being overbearing on what the user can or cannot do. For example, it would be extremely annoying if the robot's collision avoidance made it actively try to stay away from walls "because the user could potentially make a sudden turn at full speed towards the wall, and then a collision could result, so therefore we should try to avoid that scenario because, damn it, we promised collision avoidance!". The net effect would be that the user might have a very good reason for wanting to drive close to a wall, but can't because of artificial limits imposed by the software. It's better to not waste time on those kinds of 'features' and instead focus on making the core experience better. In this case, focusing on the two-way video and reducing chat lag time. Incidentally, those kinds of 'features' seem like exactly the kind of thing that a mature company would focus on (waste time on) and lose sight of making the actual product better, because it combines so many things that mature companies would like -- delivering on promises to the user, it sounds good on paper, etc. But it loses sight of the fact that the goal is to make the end product better, not to deliver a laundry list of things.

pavs 16y ago
Why does the collision avoidance has to be a hard limit? Allow it to be toggled on and off at will by the operator and it solves all problems you mentioned. Collision off on cramped places with furniture. Collision on when too many moving object (people walking around).
- robobenjie 16y ago
  Actually we allow people to hit walls, for the reasons you mention above, but we make sure that the robot slows down a little before hitting them. Basically we allow any movement but you can only go fast when you are in an open area and if the robot detects that you are going to barely clip an obstacle you barely miss it instead (thats the hope anyway) -Benjie, anybots obstacle avoidance guy
  - spoon16 16y ago
    what happens if I drive it down stairs?
    - robobenjie 16y ago
      It depends on how long the stairway is. It has a bumper around the head to protect it, but odds are you are going to need some repairs. We are using the Hokuyo Lidar sensor: http://www.hokuyo-aut.jp/02sensor/07scanner/urg_04lx_ug01.html http://www.hokuyo-aut.jp/02sensor/07scanner/urg_04lx_ug01.ht... which sees an ankle height slice of the world so stairs down are invisible to us at the moment. :/
      - tudorachim 16y ago
        Hey, I think my original post was totally misinterpreted. I was referring to the instance near the beginning where Arrington drove the robot forward, directly into Dan Casner, a few times. That mistake will almost certainly be repeated by everybody who uses QA, as they try to get a better view of the people they're talking to. So, I was commenting on the lack of a hard limit on how close one can get to objects directly in front of the robot. Will this be addressed in the final implementation? I believe there might also be a ROS package that detects ankles (well, legs) in 2D scans, which could help.