Sensors on ECVs

[cobright]

Oh man, this topic is checking off all my boxes.

In which case Disney will have cornered, and control of, the lucrative ECV rental market within their parks.
Remember, Disney does nothing unless it generates cash income, or eliminates cash outflow.

Funny thing... ECV rental is not terribly lucrative. I mean, the companies make money, but they really have to work for it. The margins are low and the volume has to be really high. They also employ a lot of people and pay them less than Disney would have to. I'm not convinced that Disney wants to start baking this pie in their own kitchen.

This also means that of they insisted on featured providers putting the sensors in, the featured providers would tell them to pound sand and Disney would loose out on whatever kickbacks they get (and make no mistake they get something for listing companies as a featured provider

kickback they get is indemnity. They most likely require a $2 million insurance rider on every scooter they facilitate as a preferred agent. Not terribly expensive, but if Disney 'cornered the market' on this they would have to cover that risk themselves.

Disney also doesn't really want to rent ECVs, otherwise they would rent them out then elves rather than leasing space to a third party company for the daily rentals

This is my thinking as well. For a whole host of reasons, I think Disney is happy letting someone else handle this very specialized product for them.

(I found this: from the DOJ "Wheelchairs, Mobility Aids, and Other Power-Driven Mobility Devices" where it clearly states that an entity does not have to allow certain types of devices because of "legitimate safety concerns"; until there is a legal precedent regarding devices equipped with these sensors, we should assume that Disney has the right and the ability to restrict devices on their property, based on safety concerns alone.)

The paper is a little more accommodating than that. For one thing, it says "Such safety requirements must be based on actual risks, not on speculation or stereotypes about a particular type of device or how it might be operated by people with disabilities using them." And the burden of proving that risk lies firmly on the business trying to restrict the mobility device. Not only that, but where actual risk exists, the business may only restrict access to that type of mobility device to the extent necessary to mitigate it. A gas powered mobility aid certainly poses a risk indoors, but if the venue is very open and not crowded then they must be accommodated in the outdoor areas. that sort of thing.

Disney *could* legally say "No electrically operated/battery powered personal mobility devices are allowed on Disney World Resort property unless they have collision avoidance sensors installed", thereby forcing *everyone* - regardless of whether they rent or own - to have those sensors installed prior to entering Disney property and using the device.

There is actually a precedent for this, I will find it, but simply reducing some risk is not enough reason if it means limiting disabled access. Just the physics of adding the weight of a mobility device to a disabled person makes them a little bit more of a risk to other pedestrians in the event of a collision. The ADA has come down several times to support the position that disabled peoples right to equal access means that, to a reasonable extent, we all bear that additional risk. I suspect that people in mobility devices aren't actually significantly more likely to cause a significant injury than anyone else. And I suspect that these brutally simple sensor devices will not reduce the overall numbers of accidents by much. They may keep the ECV from rolling into someone but the sudden unplanned stops make it more likely that someone will run into them.

Actually, they can't, there is legal precedent set here, oddly enough during the lawsuit involving people having wanted to use Segways as a mobility devices. The lawsuit stated that Segways were not commonly recognized as mobility devices, ECVs are and as such cannot be restricted.

Actually, this one's a funny case. Segways are recognized as mobility devices by the ADA (see the ADA paper mamabunny linked). In the lawsuit, BAUGHMAN v. WALT DISNEY WORLD CO, Disney produced expert testimony that the conditions of the park (crowded) meant that a segway posed significant risk to pedestrians and could not be operated without doing so. Baughman had the opportunity but did not produce any expert that could counter this. So the Segway ban stood because Disney showed in court that there was a significant risk.

And it's funny all this is so timely I just started playing with some new tech that rolled out just last week that's doing some crazy stuff for collision avoidance... Maybe I'll post my new feature road-map and put up some demonstration footage... maybe tomorrow...

 

[cmwade77]

Actually, I would really.like sensors that gave something like a tactile feedback. Not something that takes control, but feedback. For example when trying to Park the scooter at Voyages of the Little Mermaid, I ended up bumping into the wall because it is so dark and hard to see. Of course this wasn't hard enough to damage anything, but I would prefer to avoid such a thing and simple tactile feedback to let me know I am too close to the wall would have helped.

 

[mom2rtk]

Oh man, this topic is checking off all my boxes.

Funny thing... ECV rental is not terribly lucrative. I mean, the companies make money, but they really have to work for it. The margins are low and the volume has to be really high. They also employ a lot of people and pay them less than Disney would have to. I'm not convinced that Disney wants to start baking this pie in their own kitchen.


kickback they get is indemnity. They most likely require a $2 million insurance rider on every scooter they facilitate as a preferred agent. Not terribly expensive, but if Disney 'cornered the market' on this they would have to cover that risk themselves.


This is my thinking as well. For a whole host of reasons, I think Disney is happy letting someone else handle this very specialized product for them.


The paper is a little more accommodating than that. For one thing, it says "Such safety requirements must be based on actual risks, not on speculation or stereotypes about a particular type of device or how it might be operated by people with disabilities using them." And the burden of proving that risk lies firmly on the business trying to restrict the mobility device. Not only that, but where actual risk exists, the business may only restrict access to that type of mobility device to the extent necessary to mitigate it. A gas powered mobility aid certainly poses a risk indoors, but if the venue is very open and not crowded then they must be accommodated in the outdoor areas. that sort of thing.


There is actually a precedent for this, I will find it, but simply reducing some risk is not enough reason if it means limiting disabled access. Just the physics of adding the weight of a mobility device to a disabled person makes them a little bit more of a risk to other pedestrians in the event of a collision. The ADA has come down several times to support the position that disabled peoples right to equal access means that, to a reasonable extent, we all bear that additional risk. I suspect that people in mobility devices aren't actually significantly more likely to cause a significant injury than anyone else. And I suspect that these brutally simple sensor devices will not reduce the overall numbers of accidents by much. They may keep the ECV from rolling into someone but the sudden unplanned stops make it more likely that someone will run into them.


Actually, this one's a funny case. Segways are recognized as mobility devices by the ADA (see the ADA paper mamabunny linked). In the lawsuit, BAUGHMAN v. WALT DISNEY WORLD CO, Disney produced expert testimony that the conditions of the park (crowded) meant that a segway posed significant risk to pedestrians and could not be operated without doing so. Baughman had the opportunity but did not produce any expert that could counter this. So the Segway ban stood because Disney showed in court that there was a significant risk.

And it's funny all this is so timely I just started playing with some new tech that rolled out just last week that's doing some crazy stuff for collision avoidance... Maybe I'll post my new feature road-map and put up some demonstration footage... maybe tomorrow...

I'm quite sure Disney doesn't have any interest in keeping enough ECV's on hand to handle peak crowds. They are happy to be the price leader so people CAN rent from them in a bind, but also happy to let the offsite companies have excess inventories sitting around in non-peak times.

 

[cmwade77]

I'm quite sure Disney doesn't have any interest in keeping enough ECV's on hand to handle peak crowds. They are happy to be the price leader so people CAN rent from them in a bind, but also happy to let the offsite companies have excess inventories sitting around in non-peak times.

This is another reason that Gold Mobility's system of selling off their ECVs after they are 6 months old is a genius move, as I am sure they have it staggered where they are selling of 1/12 each month to be replaced, which means they can scale back in off peak times (whatever those are now) and scale up at peak times. Didn't really realise how genius that every 6 months thing is until I started looking at all of the benefits it gives them and it helps people who need to buy decent equipment at a good price as well as let's the scooter manufacturers sell more devices than they normally would have..... absolutely amazing how well it works for everyone and ensure they still turn a profit.

 

[cobright]

Actually, I would really.like sensors that gave something like a tactile feedback. Not something that takes control, but feedback. For example when trying to Park the scooter at Voyages of the Little Mermaid, I ended up bumping into the wall because it is so dark and hard to see. Of course this wasn't hard enough to damage anything, but I would prefer to avoid such a thing and simple tactile feedback to let me know I am too close to the wall would have helped.

You can let the computer do more of the driving for you if it is a smarter computer. Just putting a pair of proximity sensors on the front bumper that cut power if something gets within 2' is just courting frustration. And it's a little more limiting when dealing with a traditional ECV scooter with an actual steering wheel (yoke? handlebar?) that steers by physically turning the wheels.

The platforms I've been working with use twin motors and what's called differential steering. They steer by adjusting power and direction to each side of power chair. And with a little bit of clever they can be made to drive the chair side to side as well.

The sensor package I have on my friends Jazzy includes doppler microwave radar, lidar, infrared (like the proximity sensors on the Disney ecvs) and Computer Vision (video cameras fed to computer module).

To get you what you want I would put in a display, something with a red color scheme that doesn't bodge up your night vision or annoy other guests, that would simply plot out everything around you in relation to your ecv. So you know exactly what's there and how close, even in pitch black.

On my friends chair, I get the computer much more involved. It can get her from one end of the park to the other without much more than a few seconds of her actively controlling it. With her in full control, it will take her at any human speed she wishes (about 15mph), wherever she wants to go, through crowds (it slows down a fair bit when people are close), and with practically zero chance of hitting anything. That's the part I'm most proud of. It will carry her much faster if she needs it too, but just like an abled person, running flat out increases the risk of whapping someone.

As computers get smarter and faster and smaller we can see the complexity of computer controls improve. My favorite comparison is these quadcopter drones. You can get one now for about $1000 that will fly autonomously from point A to B to C (etc.) based on clicking points on a google map screen. Along that flight it will actively avoid moving and stationary obstacles, all the while flying at 30+mph. All that using a computer the size and weight of a Snicker Bite Size.

And just when I start feeling like I'm really getting somewhere, the computer chip company Nvidia drops a new development board that makes this kind of autonomous and semi-autonomous motion even better, easier, smarter.

I mean, sure, this is probably how SkyNet starts but ... until then I'm onboard.

 

[cmwade77]

Ok, I will admit that while what you describe sounds interesting, it is a little too far down the road toward autonomous driving for me. Seriously, how long before some outside person or group of people hack such a system to control the chair to go where they want it to, perhaps driving it I to a large body of water, etc.?

But I have similar feels about self driving cars, even the "driver assist" features like Lane control, etc. While I can see the potential good, sadly, I think the potential evil actually far outweighs any good that could come from the tech. Because I could see your ideas being very useful for those who have a hard time controlling the device for extended periods of time.

If you need further ideas of just what could go wrong, check out the episode of Doctor Who that showed ATMOS as a technology for cars. It should give you pause for thought.

 

[cobright]

Seriously, how long before some outside person or group of people hack such a system to control the chair to go where they want it to, perhaps driving it I to a large body of water, etc.?

Leaving all matters of encryption beside, supposing that there was no defense to outside interference... the main defense against this sort of attack is that there is no long distance radio link on my system. Someone with murder in their heart would have to get within 100' of my controller to be able to hack it at all. And then, they can only hack the computer system. There is still the human controls which the computer never touches. In particular, there is a big red button that fires an interlock with the main battery line running to the motors. And yet, there is still the matter of encryption to come back to. Hacking my controller would require you to be onsite, following a moving target, and crack an AES-256 key. A room full of super computers would take 3×10⁵¹ years to do this; it's not like guessing a 4 digit pin or 8 letter password.

In short, it would be far easier and cheaper to simply grab the handles of a traditional wheelchair and push them into a lake. It would be infinitely easier to kill someone almost any other way than that which you describe. And should someone try, my system would allow my friend, Aisling, to flee from her attacker like not too many people in wheelchairs can. Her chair will accelerate in seconds to an absurd speed of 50mph on a straight away. It will not allow her to make steering controls that will result in the chair tipping. It will actively do its best to avoid hitting people and things (at 15mph I can claim near 100% collision avoidance. At 50mph... who knows. It's for an emergency.).

While I can see the potential good, sadly, I think the potential evil actually far outweighs any good that could come from the tech.

Here's the potential for good. Hundreds of thousands of people exist homebound or severely limited in what they do because they are afraid they might bump into something or someone if they get an ecv or powerchair. That is a fact. My chair lets someone with moderate epilepsy leave his house because he knows that if he has a seizure his chair will probably alert him before it happens but in any event will pull itself over and park. My system would allow some of the 500k people with cerebral palsy, millions with Parkinson's, tens of thousands with ALS to remain an active participant in their lives and community.

The potential for evil is ... that someone might decide to murder a handicap person? Someone who already has the skills and equipment necessary to crack a global bank or NASA?

 

[cmwade77]

Actually, the issue isn't so much your average person hacking the system, as the government doing so in order to control where people can or can't go and you can bet that the government would love to find a way to prevent people who are on disability from ever leaving their house, let alone going place like Disney, etc. Seriously, they think a disabled person should have to stay in their house all day every day if they are on disability. Don't believe me? Look at various policies for those on disability and notice how many are geared towards keeping them at home. In fact there are certain situations where if the person even goes outside, they automatically loose their disability payments. I think the policies are wrong and need to be corrected, luckily the person I knew with the issue of not being able to go.outside was finally able to get a court to rule against that policy, but it is what the government wants to do. And there are people who could (and want to) get off of disability, if insurance would help them get an ECV or wheelchair, but they can't afford to buy one on their own due to how little disability pays and insurance will only pay if it is needed to get around the house. They won't give them the tools they need to no longer be dependent on the government.

That's the rub is your system would simultaneously make it possible for these people to leave the house while allowing the government (who would have the resources to hack it) to prevent them from doing so. As I said, as much potential as there is for good, which you rightfully point out, there is more potential for evil. But if you truly can overcome that in a way that not even the government with their virtually infinite resources can crack into, you might be onto something.

Also, encryption isn't what is used to be, thanks to cloud computing, you could setup 100×10⁵¹ super computers in a cluster and crack most encryption in the matter of a couple miliseconds. This would be cost prohibitive for most people of course, but again someone with the resources the government has could do it.

Please note that I am not trying to downplay the amount of good your system could do, as I feel that is absolutely incredible, but I am stating that I think you are grossly underestimating the amount of potential evil there is that will find way to abuse the technology and it will happen. There unfortunately are a lot of forms that evil takes out there these days, murder was but one example.

Of course I suppose if you truly have a mechanical (not digital, but physically mechanical) off switch for the features that reverts to full manual control in the event that this happened, then any such hacks would automatically be overridden, which could work, but that switch would truly need to be physical.

 

[iuki]

Yes, it does sound more problematic than helpful, especially since, when there are crowds (like always) there are always people close by.

I have my personal ECV. People jump in front of me dragging their children behind them. I am constantly trying to avoid hitting people cutting in front of me. Having sensors are not practical in crowded situations.

 

[cobright]

That's the rub is your system would simultaneously make it possible for these people to leave the house while allowing the government (who would have the resources to hack it) to prevent them from doing so.

To state it again, whoever decides to attack someone in a smart wheelchair will be engaging in the most idiotic criminal enterprise. There is literally dozens of easier and cheaper ways to achieve the same goal, some of them might not even be illegal. It would literally be easier and much cheaper for a government to, having decided that handicap Bob had to go, send in a wetworks assassin or fire a missile or pay a 'drunk driver' to hit him as he crosses the road. It would be easier for them to hack his pharmacy's Parata rx dispensing machine to give him a similarly shaped alpha channel blocker. These are all things that are theoretically (at least) possible now and all of them would be more cost effective than trying to take control of wheelchairs. If this risk is what keeps us from doing new things and caring for people, then why get out of bed anymore at all?

On a more optimistic note (and if anyone cares). I cobbled together some demo footage for a new computer module and software I just started playing with. A brief rundown of whats involved in a semi-autonomous powerchair control computer. It is a computer that takes the input from a variety of sensors and then decides, on its own, to alter the controls of the wheelchair. In the simplest sense, the sensors described in the OP of this thread do this. My system is essentially the flight control computer of one of those quad-copter drones blown up to handle the larger motors of a wheelchair. The technology is very close to the same.

Or it was until about 2 years ago. that's when computer vision (CV) began to really pick up steam. Intel offered a CV processing board that made adding video camera input to my system a lot easier. Now the computer can identify what is around the wheelchair, whether it is people, animals, stationary objects, etc. This lets my computer work a lot faster because it doesn't have to spend time tracking things it now identifies as stationary. It also lets me do something neat, it lets the computer tell who, among the people around it have looked at the wheelchair. Who has made eye contact with it, so to speak. This is something I was just beginning to use to let the program decide if a particular person has seen and recognized the wheelchair (and operator of course) for what they are, at which points it can make a few assumptions about their behavior. For example, someone who demonstrates that they have seen and identify the wheelchair are less likely to walk in front of it, so the computer can check in on where that person is less often; maybe 10 times a second instead of 300.

And then, like 3 weeks ago, Christmas came early. Nvidia dropped a smaller, cheaper version of its Jetson board, with built in machine learning and computer vision hardware.

• Here's a clip from a demo that shows the motion estimation software at work -
• Here's a demo of object detection running in real time
  The Jetson hardware lets me process 8 video streams like this.
• Another object detection demo done at the pace of a James Bond opener
• An autonomous drone, flying a 750m winding trail without any human intervention
• An interesting explanation of something I'm very excited about https://news.developer.nvidia.com/teaching-cameras-to-read-body-language-with-ai/

Check that last part out, somewhere in that page you'll find this quote, "helping autonomous vehicles read human intent". Now instead of simply identifying an object as human, the machine can observe how they move and learn from that how they are likely to behave. This is something people do subconsciously, but we now have a way to make this technique useful at a machine level. Human minds see a person walking a certain way and we know that person is in a hurry, or they're drunk, or whatever. Sometimes we notice that a person is always looking in the direction of the place he or she intends to go. All important information to a machine that wants to get through a crowd of these people without hitting anyone and, just as important, without slowing its rider down to an inhumanly slow pace.

 

[Starwind]

Here's the potential for good. Hundreds of thousands of people exist homebound or severely limited in what they do because they are afraid they might bump into something or someone if they get an ecv or powerchair. That is a fact. My chair lets someone with moderate epilepsy leave his house because he knows that if he has a seizure his chair will probably alert him before it happens but in any event will pull itself over and park. My system would allow some of the 500k people with cerebral palsy, millions with Parkinson's, tens of thousands with ALS to remain an active participant in their lives and community.

Cobright, I just wanted to say that I think the work you are doing is really cool an the potential it has is tremendous. I also like that you have recognized security concerns and build some protections into the system -- so many medical and IOT devices DON'T do that, lacking encryption, using hard coded default passwords [if any password at all!], etc. it is pretty scary.

Keep up the good work !

As for the ECV sensors Disney is testing, I think whoever came up with the idea perhaps hasn't actualy spent much time **using an ECV in the parks**. Like really spending a good day or more, in the crowds, seeing what it is like. As another poster noted, one is constantly dealing with people coming into your path -- nearly running into you. The biggest hazard is not the ECV running into things, most of the time IME it is people running or nearly so into the ECV. Their sensors as deployed sound like they create more problems than they solve.

SW

 

[lookin4goofy]

I just rented an ECV at MK last week. The safety feature was dangerous and annoying. My scooter would stop if it sensed anything on either front side. The park was very busy and the scooter stopped constantly. Then it would start back up VERY slowly, just creeping along for awhile. It would stop for people, but also for walls and poles. Taking the scooter through lines was a test of patience, mine and everyone else around me. Everywhere we walked, it would stop abruptly if it sensed anything. So that was many stops in every line as it sensed poles and walls. I felt like at any time I could get run into from behind and it had to be terribly annoying for people behind me. Sometimes when the scooter was in the creeping stage prior to resuming full speed, very nice people would step aside and motion me to go ahead. Well, I couldn't, because the darn thing wouldn't move. I kept explaining to people so they wouldn't want to harm me. Just kidding, kinda. The strangest thing was that it seemed to only sense things and people from the front sides. I could run into something straight ahead full speed. Full speed on MK scooter rentals is much slower than on outside rentals, but still. I hated it and would always get an outside rental in the future if possible.

 

[cmwade77]

I just rented an ECV at MK last week. The safety feature was dangerous and annoying. My scooter would stop if it sensed anything on either front side. The park was very busy and the scooter stopped constantly. Then it would start back up VERY slowly, just creeping along for awhile. It would stop for people, but also for walls and poles. Taking the scooter through lines was a test of patience, mine and everyone else around me. Everywhere we walked, it would stop abruptly if it sensed anything. So that was many stops in every line as it sensed poles and walls. I felt like at any time I could get run into from behind and it had to be terribly annoying for people behind me. Sometimes when the scooter was in the creeping stage prior to resuming full speed, very nice people would step aside and motion me to go ahead. Well, I couldn't, because the darn thing wouldn't move. I kept explaining to people so they wouldn't want to harm me. Just kidding, kinda. The strangest thing was that it seemed to only sense things and people from the front sides. I could run into something straight ahead full speed. Full speed on MK scooter rentals is much slower than on outside rentals, but still. I hated it and would always get an outside rental in the future if possible.

Yikes! Did you call the number that they sat to call if you are having problems in the park? Seriously, I think that is the only way this will get fixed.

 

[lookin4goofy]

I talked to the cast member at rental location. They consider it a safety feature, not a problem.

 

[cmwade77]

I talked to the cast member at rental location. They consider it a safety feature, not a problem.

The problem is you did at the rental location, no, call them out to the middle of the park and show them the ECV won't move. Don't try to allow people around you if you are in a queue either, show them what is really happening.

 

[Tallis]

This sensor things on ECVs are nothing short of a nightmare at Disney. I own my old older scooter, with no sensor. It was broke down in February, when we went to WDW for a couple of days, so I borrowed my Dad's modern scooter. I didn't know about the sensors. I thought it kept breaking down on me. It would stop dead it in tracks, for apparently no reason, then refuse to move at all for a few minutes. I know I made many people angry during those two days. Everyone thought I was just driving like a jerk and stopping for no reason. It wasn't until I got home that I found out about the "safety sensors," after telling my dad the scooter needed fixing. Never again.

I am only thankful that I will be out of this whole scooter situation soon. Had my first knee replacement 4 weeks ago, and having the other in July. I can't wait to walk the parks again without dealing with any scooter. I feel horrible for people that are stuck with them, especially hearing this new development. Over 5 years on a scooter and I've never injured myself, or anyone else, until I used this scooter with so-called safety features built in. This is not going to go well, from a safety perspective.

 

[Betty Rohrer]

The problem is you did at the rental location, no, call them out to the middle of the park and show them the ECV won't move. Don't try to allow people around you if you are in a queue either, show them what is really happening.

I would add have a CM watching the line call for you or have a ride lead or what ever they are called at Disney to see what is happening in line causing a problem in his/her line

 

[mamabunny]

:::ahem:::

I wonder what would happen if the Parks units that are equipped with sensors had a Post-it Note "accidentally" dropped over the sensor?

Is it like when we put a Post-It over the auto-flushing toilets in the bathroom? Would it stop the scooter completely? Or is there an override that would then allow it to function "normally" without the sensor interfering?

This is all just conjecture, mind you...

...Unless someone wants to try it, and report back!

 

[cobright]

The biggest hazard is not the ECV running into things, most of the time IME it is people running or nearly so into the ECV. Their sensors as deployed sound like they create more problems than they solve.

For most ECV/Wheelchair users I think you are dead on here. I have sensor data and video from two test cases covering hundreds of hours of driving their powerchair through crowded environments (often the WDW parks) and the hardest thing for me has been dealing with the fact that even a big 200lb powerchair is, as often as not, invisible to the average pedestrian. I actually have the computer vision system make a distinction between people who have looked at the wheelchair (and of course its driver) and those who have made 'eye-contact'.

Kinda funny, in a nerdy way... The first camera system I used was a rotary camera, it would spin to take a 360deg photo. The camera rides above the head of the chair's driver, about at eye level with the average pedestrian. At first I had it running all the time, but the power drain was an issue, so I had it spin down when the vehicle went idle and spin back up when the controls were engaged. All of a sudden my 'eye-contact' data shot through the roof. It turns out, the camera makes a little chuckling sound when it spins up. It's sort of like a soft clearing of throat. A sort of non-verbal "ahem" as one might use when walking through a crowd. I don't believe in "Eureka" moments, but this one comes about as close as I can find. I now use a pair of solid state (silent) 3D 360deg cameras which can run all the time and have no moving parts, but I installed a completely unnecessary housing where the old camera used to sit with a little speaker inside that will play a recorded version of that "chuckle" whenever someone is looking at it but not making eye-contact.

I also like that you have recognized security concerns and build some protections into the system -- so many medical and IOT devices DON'T do that, lacking encryption

Yes, IoT got a real bloody nose not too long ago, and rightfully so. And I will be the first to admit, a huge part of some engineer's job will be chasing down all of the vulnerabilities I am building into my system. I am not a software engineer. But the risk of one of my one-off prototypes being compromised in any meaningful way are pretty remote, and anything novel I create will not be sold as a complete system anyway but patented and sold off as individual items. The actual coding thrown out and rewritten by people who know how.

I just rented an ECV at MK last week. The safety feature was dangerous and annoying. My scooter would stop if it sensed anything on either front side. The park was very busy and the scooter stopped constantly.

And this is the most maddening aspect of this whole story. Consider the flow of people in a crowded segment of park as a whole, the worst thing you can possibly do is make one or more of the individual elements of this scene act unpredictably. People subconsciously identify objects and track motion and route plot themselves through their environment and all of that is based on some assumptions of how these objects in motion will behave. If you want to see chaos, just stop at the top of an escalator. Or perhaps stop your ecv at a Disney park.

Full speed on MK scooter rentals is much slower than on outside rentals, but still.

Top speed of MK scooters is criminally slow. Mobility tech is supposed to bridge the gap between what a handicapped person is capable of and what an average person is capable of. Perfect Mobility tech matches the one to the other. An average able adult can run 15mph over short distances. I haven't seen the onsite rental scooters do more than a brisk jog or maby race-walk speed.

I wonder what would happen if the Parks units that are equipped with sensors had a Post-it Note "accidentally" dropped over the sensor?

I would like to see a picture of the sensor. Maybe time for another trip to the park. But it depends really on what type of sensor is being used. I strongly suspect it is a basic (industrial) infrared proximity sensor. It could also be an ultrasonic "sonar" sensor but those aren't usually susceptable to getting dirty and needing to be cleaned as an earlier post mentioned being told about these scooters. It sounds like getting dirty caused the sensor to read a false positive, which means covering them will do the same thing.

This is a bit different that most auto flush sensors. Those (again, usually) measure changes in the sensor response. the sensor works by emitting an infrared light and also has an IR receiver that measures how much, if any, IR light gets reflected back. This is how most proximity sensors work. But because lots of things affect how much IR that reciever detects (changes in ambient light, temperature, just lots of things) the flush valves can't just set a threshold at which point it decides that a person is there or not, this would make them likely (as early models did) to go off all the time for no reason (something that still happens on failing units). What it does is take lots of readings and sets a baseline for unoccupied. You can see it flashing each time it gets a threshold reading and stops when you move away from it. So blocking the sensor here, in the first part tells it that someone is in front of it and it should not flush, much the same as putting it infront of an ECV sensor would tell it not to allow the scooter to drive. But on the toilet valve it goes further by, over time, ruining the baseline to the point that after a few minutes even taking off the postit note would not restore normal function, for a little while anyway.

Now what does exist on every proximity sensor I can imagine being used here is a sensitivity adjustment. A tiny flat head screw used to dial in how close something gets before the sensor closes the switch. So... a tiny eyeglass repair screwdriver and a little privacy...

 

[serenitynow]

For most ECV/Wheelchair users I think you are dead on here. I have sensor data and video from two test cases covering hundreds of hours of driving their powerchair through crowded environments (often the WDW parks) and the hardest thing for me has been dealing with the fact that even a big 200lb powerchair is, as often as not, invisible to the average pedestrian. I actually have the computer vision system make a distinction between people who have looked at the wheelchair (and of course its driver) and those who have made 'eye-contact'.

Kinda funny, in a nerdy way... The first camera system I used was a rotary camera, it would spin to take a 360deg photo. The camera rides above the head of the chair's driver, about at eye level with the average pedestrian. At first I had it running all the time, but the power drain was an issue, so I had it spin down when the vehicle went idle and spin back up when the controls were engaged. All of a sudden my 'eye-contact' data shot through the roof. It turns out, the camera makes a little chuckling sound when it spins up. It's sort of like a soft clearing of throat. A sort of non-verbal "ahem" as one might use when walking through a crowd. I don't believe in "Eureka" moments, but this one comes about as close as I can find. I now use a pair of solid state (silent) 3D 360deg cameras which can run all the time and have no moving parts, but I installed a completely unnecessary housing where the old camera used to sit with a little speaker inside that will play a recorded version of that "chuckle" whenever someone is looking at it but not making eye-contact.


Yes, IoT got a real bloody nose not too long ago, and rightfully so. And I will be the first to admit, a huge part of some engineer's job will be chasing down all of the vulnerabilities I am building into my system. I am not a software engineer. But the risk of one of my one-off prototypes being compromised in any meaningful way are pretty remote, and anything novel I create will not be sold as a complete system anyway but patented and sold off as individual items. The actual coding thrown out and rewritten by people who know how.


And this is the most maddening aspect of this whole story. Consider the flow of people in a crowded segment of park as a whole, the worst thing you can possibly do is make one or more of the individual elements of this scene act unpredictably. People subconsciously identify objects and track motion and route plot themselves through their environment and all of that is based on some assumptions of how these objects in motion will behave. If you want to see chaos, just stop at the top of an escalator. Or perhaps stop your ecv at a Disney park.


Top speed of MK scooters is criminally slow. Mobility tech is supposed to bridge the gap between what a handicapped person is capable of and what an average person is capable of. Perfect Mobility tech matches the one to the other. An average able adult can run 15mph over short distances. I haven't seen the onsite rental scooters do more than a brisk jog or maby race-walk speed.



I would like to see a picture of the sensor. Maybe time for another trip to the park. But it depends really on what type of sensor is being used. I strongly suspect it is a basic (industrial) infrared proximity sensor. It could also be an ultrasonic "sonar" sensor but those aren't usually susceptable to getting dirty and needing to be cleaned as an earlier post mentioned being told about these scooters. It sounds like getting dirty caused the sensor to read a false positive, which means covering them will do the same thing.

This is a bit different that most auto flush sensors. Those (again, usually) measure changes in the sensor response. the sensor works by emitting an infrared light and also has an IR receiver that measures how much, if any, IR light gets reflected back. This is how most proximity sensors work. But because lots of things affect how much IR that reciever detects (changes in ambient light, temperature, just lots of things) the flush valves can't just set a threshold at which point it decides that a person is there or not, this would make them likely (as early models did) to go off all the time for no reason (something that still happens on failing units). What it does is take lots of readings and sets a baseline for unoccupied. You can see it flashing each time it gets a threshold reading and stops when you move away from it. So blocking the sensor here, in the first part tells it that someone is in front of it and it should not flush, much the same as putting it infront of an ECV sensor would tell it not to allow the scooter to drive. But on the toilet valve it goes further by, over time, ruining the baseline to the point that after a few minutes even taking off the postit note would not restore normal function, for a little while anyway.

Now what does exist on every proximity sensor I can imagine being used here is a sensitivity adjustment. A tiny flat head screw used to dial in how close something gets before the sensor closes the switch. So... a tiny eyeglass repair screwdriver and a little privacy...

Thank you! I wish that everyone who designed public spaces would spend a week in a wheelchair, power chair and ECV. It is difficult to navigate an ECV or wheelchair in an environment where other are walking, gawking, quickly stopping and generally acting like typical distracted human beings in a congested environment. The sensors are only increasing this problem for ECV users. One can’t suddenly stop short in a crowd without creating chaos behind them.

I have been on multiple trips to Disney using an ECV - on each trip at least one person runs into ME. As soon as I enter the parks I become invisible to many people. I have even had the experience of being parked, not moving, against the wall outside of a restroom and tripped over.

While the ECV is moving, people constantly dart out in front of the ECV in their hurry to get to the next ride or whatever. People also follow too closely behind so that when someone cuts in front of the ECV, if the device slows or stops, people behind the ECV run into you. I predict more guests are going to be injured who are behind the ECV with sensors that keep stopping and starting the device.

It’s so much worse now that people walk around looking at their phones. I think it’s because a person using a wheelchair are a little below their line of sight while walking and texting.

Slowing down the speed of the ECV might have been a good idea on paper, but is generally useless as a safety measure in a crowded amusement park. There is no way to speed through the congestion at Disney on a ECV, it’s slow turtle speed all the way. My family has to slow down their walk if they want to stay with me in the parks. Walking humans can weave in and out of crowds, “rolling” guests are just stuck creeping along behind the crowds.

Thank you for the work you are doing to help all of us “share the road”!