Shop OBEX P1 Docs P2 Docs Learn Events
Ping....how reliable at higher speeds? — Parallax Forums

Ping....how reliable at higher speeds?

Bulslayer85Bulslayer85 Posts: 34
edited 2012-04-20 18:26 in Accessories
K so my robot is big essentially, like 500 pound motorcycle big, think KITT only a motorcycle and bare with me. Prob is i need object detection sensors suitable at higher end speeds of 60 mph or more that still return good signals of an object at a location. Now Ping or ultrasonic senors seem most likely to be used sense they can work threw fog and hopefully rain better than say an IR could. My question is "are they reliable at those higher speeds the bike will be going?"

See I am trying to develop a system that monitors where cars are if they are close to the bike. The idea is that if the rider tries to move into a lane that has a car he or she cant see then the bikes computer will pick it up and warn the rider that the lane is occupied. As an experienced motorcycle rider i can say its hard to just turn your head and look because every second your head is turned is every second your bike will drift in that direction, basic motorcycle physics "you go where you look" and some people have a harder time judging distance than others.

Comments

  • Mike GreenMike Green Posts: 23,101
    edited 2012-04-19 09:34
    If you look at the documentation for the PING, you'll see that it can make at most 50 measurements per second at maximum range which is only about 10 feet. That sounds like something completely inappropriate for your needs. Fog and rain will tend to absorb / reflect some of the ultrasonic energy as well.

    You may need some kind of radar device to detect cars (large metal masses) nearby.
  • Bulslayer85Bulslayer85 Posts: 34
    edited 2012-04-19 11:25
    Well actually if they measure accurately up to 10ft then they are good range wise as im only looking for vehicles with in close proximity to the bike. A quick glance to the right or left will tell me if the cars are right there, i need the pings to tell me whats behind me to the point where turning my head to look could be considered dangerous at high speed. so they would sit towards the back of the bike and angle out to the right and left of bike to find vehicles in next lanes. I meant to ask if their accuracy drops at higher speeds of 60mph or more because the bike is moving faster than they can return a signal. But either way you are probably correct in saying a Radar would be more appropriate which it would be but they are expensive and i have a limited budget to prove this concept on.
  • Bulslayer85Bulslayer85 Posts: 34
    edited 2012-04-19 11:35
    After reading the documentation I know the parallax ultrasonic s wont work cause they not outdoorsy, so question is are there outdoor ones that will return a signal fast enough to catch up with a vehicle moving 60 +mph and at a cheap price? So would something like http://www.futurlec.com/Ultrasonic_Sensors.shtml work
  • Mike GreenMike Green Posts: 23,101
    edited 2012-04-19 12:07
    There's no way I'd trust a simple ultrasonic sensor with something as dangerous as adjacent cars at speeds like 60MPH. Sorry.

    The waterproof sensors you showed at Futurlec are the transducers only ... no electronics. There are some DIY electronics designs on the web, but I don't have any links to them. Some web searching will find them, but I have no idea how well they work.
  • PublisonPublison Posts: 12,366
    edited 2012-04-19 12:33
    Parallax sells weatheproof transducer, 40 Khz:

    http://www.parallax.com/StoreSearchResults/tabid/768/txtSearch/Transducer/List/0/SortField/4/ProductID/755/Default.aspx

    Y    ou would have to design the excitation circuit.

    I agree with Mike. This is not a part to be relied on for safety, otherwise ever auto manufacturer would be be using them for vehicle avoidence.

    Jim
  • W9GFOW9GFO Posts: 4,010
    edited 2012-04-19 17:59
    The Ford Flex that we rented for the trip to the Parallax Expo had blind spot sensors that do just what you describe. I did not look to see what they were using for sensors. Their range seemed to be more like 20 feet. Anytime a vehicle was in the blind spot a small light in the appropriate side mirror would illuminate.
  • GordonMcCombGordonMcComb Posts: 3,366
    edited 2012-04-20 07:50
    Ford's system uses either radar or a blob detection camera. Ultrasonic is okay for very low speed (no wind) such as that for backup sensors, but at freeway speeds ultrasonic has a lot of issues. All the side and front view sensors I've seen are either radar, laser, or vision. For short range vision is the ticket. The cameras are cheap and it's all done in software. This is what many Volvo models use.

    -- Gordon
  • xanaduxanadu Posts: 3,347
    edited 2012-04-20 13:53
    You can see the blind spot detectors on most cars under the side view mirrors facing back, looks like a little camera.
  • xanaduxanadu Posts: 3,347
    edited 2012-04-20 14:26
    You could just put regular cameras on the blind spots and show it somewhere. The rider should be making all of the decisions though. Being able to control all of the buttons and switches on the handlebars and gauge cluster without using fingers or turning your head would be nice. If you want to use a Ping on a bike point it at the rider and based on the riders position different things can happen, that would be sweet. It also wouldn't be reliable, you know it would get covered in insects in a 10 minute ride.
  • Bulslayer85Bulslayer85 Posts: 34
    edited 2012-04-20 18:26
    Yeah i know the Fords use the blind spot stuff that is actually where i got the idea from to do it on a motorcycle actually, Sounds like ultrasonic sensors wont work so i will see about doing some kind of radar instead. Trouble is where do i mount it lol.

    I know this seems all redundant cause this is what rear view mirrors are for but i plan on making a motorcycle version of KITT, well ok sorta, the AI wont be that cool but the idea is the bike can operate its self and for that it needs to know where everything is.
Sign In or Register to comment.