Wooden compass with single red arrow leads people with dementia to their homes(designboom.com)
designboom.com
Wooden compass with single red arrow leads people with dementia to their homes
https://www.designboom.com/technology/wooden-compass-single-red-arrow-leads-people-with-dementia-homes-aumens-12-28-2025/
35 comments
The magnetic sensor in your phone is actually pretty weak and requires regular calibration to pick out the true magnetic field lines from all the noise.
Your phone has many magnets inside, and structures that can be passively magnetized. Your local magnetic environment changes constantly.
It's a fundamental limitation of the technology, unfortunately. Generally we use GPS location as you move to infer direction and feed that back into the compass routines. It's not great but it does mostly work sometimes
Your phone has many magnets inside, and structures that can be passively magnetized. Your local magnetic environment changes constantly.
It's a fundamental limitation of the technology, unfortunately. Generally we use GPS location as you move to infer direction and feed that back into the compass routines. It's not great but it does mostly work sometimes
Have you done orienting in the past?
One of the big tricks is to realize that magnetic north (where your compass points) is usually not the same as geographic north (where maps are drawn). The adjustment is local; here, I think magnetic north is 7 degrees off from north.
In the US, the USGS topographic maps contain the required adjustment for the covered area. Not sure about Italy.
One of the big tricks is to realize that magnetic north (where your compass points) is usually not the same as geographic north (where maps are drawn). The adjustment is local; here, I think magnetic north is 7 degrees off from north.
In the US, the USGS topographic maps contain the required adjustment for the covered area. Not sure about Italy.
I don't think that's it. For me sometimes my phone compass is (roughly) right, but sometimes points in a completely wrong direction, for example 90 degrees off. It calibrates after a while when I start walking but I guess this is what OP had in mind, not a few degree difference.
iPhones can regain the compass fix if you move the phone along a figure eight path while holding it flat and level. Worked pretty well for me in the past.
I'm fairly often doing figure-8 with my phone in front of myself to realign the compass.
Thoughtful concept but agreed. I.i.r.c the phone can assess the inaccuracy of the compass and prompt the figure-8 movement. But that is not practical here.
I like this, but of course the location it points to should be programmable at any time.
Then it's useful for everyone without dementia as well, and a bigger market is better.
For me, this is the perfect motorcycle navigation. In fact, I use an app that turns my smart watch into exactly this product--an arrow pointing to my destination at all times, no other information (distance might be nice though)
Actually, if it's for people with dementia, I think it should have a couple words on it. "Home", an actual arrow, and a distance. Two of those could be static additions, if the designer will allow their minimalism to be sacrificed for the actual goal of the product.
Then it's useful for everyone without dementia as well, and a bigger market is better.
For me, this is the perfect motorcycle navigation. In fact, I use an app that turns my smart watch into exactly this product--an arrow pointing to my destination at all times, no other information (distance might be nice though)
Actually, if it's for people with dementia, I think it should have a couple words on it. "Home", an actual arrow, and a distance. Two of those could be static additions, if the designer will allow their minimalism to be sacrificed for the actual goal of the product.
> motorcycle navigation
Have you seen the Beeline app/hardware? Pretty much a similar concept of 'follow the arrow' to get to where you need to go. https://beeline.co/
> programmable at any time
I assume a caregiver application could reprogram the location pushpin as a feature.
Have you seen the Beeline app/hardware? Pretty much a similar concept of 'follow the arrow' to get to where you need to go. https://beeline.co/
> programmable at any time
I assume a caregiver application could reprogram the location pushpin as a feature.
I hadn't seen Beeline, but it looks sweet. I may pick one up.
I always wonder if I'm just buying future ewaste, though. Looks like it requires their app, which requires their live servers. One acquisition away from end-of-lifed.
I always wonder if I'm just buying future ewaste, though. Looks like it requires their app, which requires their live servers. One acquisition away from end-of-lifed.
The Beeline app lets you start navigation and 'simulate' a Beeline on screen, if you'd like to try one.
Great, I wish you all well looking for this device — in addition to their keys — for two hours every couple of days.
After several weeks I once found the lost TV remote in the Scrabble box. However it was short lived: we had to replace it at least another five times over the next couple of years.
After several weeks I once found the lost TV remote in the Scrabble box. However it was short lived: we had to replace it at least another five times over the next couple of years.
Ah good point. I think the solution to this is simple. A second, smaller compass that points towards the first one
Something something something built in airtag-like finder?
Remembering how ”easy” it is to navigate with such an arrow in GTA, I would expect this to fail in unimaginable ways.
It could point at a river instead of the bridge over it.
It could point at an airport instead of the route around it.
It could point at a cul-de-sac, a dangerous neighborhood, a construction site, etc.
It could point at a river instead of the bridge over it.
It could point at an airport instead of the route around it.
It could point at a cul-de-sac, a dangerous neighborhood, a construction site, etc.
That depends on the implementation. It could use GPS and map to trace walkable path to destination and point the needle in the direction that you need to move towards to follow the path.
Except this doesn’t work on smartphones. Have you ever tried using GPS? It doesn’t understand what direction you’re facing unless you’re steadily moving forward.
It's because maps apps don't trust builtin compass. Which is reasonable since it fairly often loses calibration.
That's because GPS is not directional. It's a positioning system. Direction can only be inferred as the difference between locations.
GPS fundamentally has no way of determining your direction. It was never designed for anything other than giving you a single fixed 3D location on the Earth's surface.
GPS fundamentally has no way of determining your direction. It was never designed for anything other than giving you a single fixed 3D location on the Earth's surface.
Rubbish: from first principles it is obvious that direction can be calculated. You are correct for phones, which have omnidirectional antennas that can't calculate direction (also built up areas with multipath would make direction finding difficult).
Marine navigation systems are the most familiar example. Many boat GPS units advertise “GPS compass” or “true heading at zero speed.” . The antennas contain two GNSS elements, fixed a known distance apart, and the direction is found using phase comparison: they look at the tiny difference in arrival time of the same satellite signal at the two separate antennas.
You just basically confirmed what the previous commenter said. GPS is not directional. You need two of them placed some significant distance apart (think ship length, not smartphone length).
When you have two points determined with high enough precision you obviously have a direction and even orientation.
Your comment could have been just that. Useful information many people could find interesting. Instead you decided to sprinkle it with bile.
When you have two points determined with high enough precision you obviously have a direction and even orientation.
Your comment could have been just that. Useful information many people could find interesting. Instead you decided to sprinkle it with bile.
You are doubling down - I do find your misinformation to be interesting.
Have a look at images of the antennas:
https://duckduckgo.com/?q=%22GPS+compass%22+antenna&ia=image...
I do apologize for liking brutal facts - I am an engineer type.
I clicked the Furuno antenna since it looked sexy - blurb:
Calculating phase difference only needs a single satellite visible, so I would guess directionality can be precise in built up areas even where location is harder to fix. although I'm assuming multipath can be filtered out.
Have a look at images of the antennas:
https://duckduckgo.com/?q=%22GPS+compass%22+antenna&ia=image...
I do apologize for liking brutal facts - I am an engineer type.
I clicked the Furuno antenna since it looked sexy - blurb:
SC-130 features a Tri-sensor antenna that provides a high system accuracy for the heading of your vessel [for Autopilot, berthing etcetera.]. 0.25°rms
Calculating phase difference only needs a single satellite visible, so I would guess directionality can be precise in built up areas even where location is harder to fix. although I'm assuming multipath can be filtered out.
SC-130 has distance between the antennas on the order of 1 meter and has three of them (which probably let's them be closer than in two antenna unit).
If you test phase shift to one satelite you might get good accurancy for some directions and no accurancy at all for the orthogonal directions.
So maybe not ship-length but still 1 meter sized device is way too large to put under the hood of a car let wlon in a smartphone.
Again, GPS has no provisions for orientation detection. You can hack it with few antennas significantly spaced apart (thank you again for pointing out those solutions). I think it might be easier to detect location with GPS and orientation by testing phase shift to the closest cell tower (when on land). There's nothing in GPS system thst helps with determining orientation.
And you still can't drop arrogance which gets tiresome.
If you test phase shift to one satelite you might get good accurancy for some directions and no accurancy at all for the orthogonal directions.
So maybe not ship-length but still 1 meter sized device is way too large to put under the hood of a car let wlon in a smartphone.
Again, GPS has no provisions for orientation detection. You can hack it with few antennas significantly spaced apart (thank you again for pointing out those solutions). I think it might be easier to detect location with GPS and orientation by testing phase shift to the closest cell tower (when on land). There's nothing in GPS system thst helps with determining orientation.
And you still can't drop arrogance which gets tiresome.
Yes, I was being rude, but the OP and you are and choosing to write incorrect information. You in particular are redoubling your errors and arrogantly stating incorrect "facts". Here's a dual antenna direction finder smaller than a cellphone:
I had chosen the 1m Furuno because it was pretty, not because it was a great example.
I personally suspect that the people that designed the GPS system deeply cared about using the system to calculate heading information for missiles and fighters. Unfortunately it is hard to get good information about either the system design or missile antennas. I did a quick search for information about missile antennas and learnt some stuff but I didn't find anything juicy.
Disclosure: I'm no expert on any of this. I just like hacking.
Vectornav VN-310 embedded. Size: 31 x 31 x 11 mm. Power:<1.6W Weight: <15g.
I had chosen the 1m Furuno because it was pretty, not because it was a great example.
I personally suspect that the people that designed the GPS system deeply cared about using the system to calculate heading information for missiles and fighters. Unfortunately it is hard to get good information about either the system design or missile antennas. I did a quick search for information about missile antennas and learnt some stuff but I didn't find anything juicy.
Some early and experimental GPS compasses used a rotating directional antenna. The idea was to use a patch or helix antenna that has a gain pattern. Spin it, watch signal strength from satellites rise and fall, and infer orientation. A few marine and military prototypes in the 1990s did this. Once dual-antenna phase comparison became practical, that approach died.
Disclosure: I'm no expert on any of this. I just like hacking.
> vn-310
Interesting, however I'm doubtful about how this unit achieves compass function. It's filled to the brim with other sensors. It has gyros, accelerometers and magnetometers. I suspect that GNSS-compass is more of a marketing term for this product that integrates input from all sensors and keeps it over time for resting direction estimate.
On https://www.vectornav.com/products/detail/vn-310 they state accuracy as:
0.15° - 0.3° GNSS-Compass Heading (1m)
What is (1m)?
And in the brochure https://www.vectornav.com/docs/default-source/product-brochu... the model that advertises "GNSS-Compass Heading" has clearly visible two sockets for external antennas which I have no idea how far apart should be placed.
Interesting, however I'm doubtful about how this unit achieves compass function. It's filled to the brim with other sensors. It has gyros, accelerometers and magnetometers. I suspect that GNSS-compass is more of a marketing term for this product that integrates input from all sensors and keeps it over time for resting direction estimate.
On https://www.vectornav.com/products/detail/vn-310 they state accuracy as:
0.15° - 0.3° GNSS-Compass Heading (1m)
What is (1m)?
And in the brochure https://www.vectornav.com/docs/default-source/product-brochu... the model that advertises "GNSS-Compass Heading" has clearly visible two sockets for external antennas which I have no idea how far apart should be placed.
> however I'm doubtful about how this unit achieves compass function
There's your problem: why continue to deny simple facts when you're shown them so clearly? Why should anyone ever waste their time to learn you?
> It's filled to the brim with other sensors. It has gyros, accelerometers and magnetometers
I'm guessing you don't realize how small those surface mount components are. Yet you must know your phone also contains them. The first duckduckgo result for a gyroscope was the A3G4250D in an LGA-16 package 4x4x1.1 mm. I didn't look for smaller ;)
Perhaps your employment doesn't have much to do with engineering? Facts are so awkward. I despise them!
> What is (1m)?
I've found AI too be pretty good at answering similar questions - so long as you form your prompts well.
There's your problem: why continue to deny simple facts when you're shown them so clearly? Why should anyone ever waste their time to learn you?
> It's filled to the brim with other sensors. It has gyros, accelerometers and magnetometers
I'm guessing you don't realize how small those surface mount components are. Yet you must know your phone also contains them. The first duckduckgo result for a gyroscope was the A3G4250D in an LGA-16 package 4x4x1.1 mm. I didn't look for smaller ;)
Perhaps your employment doesn't have much to do with engineering? Facts are so awkward. I despise them!
> What is (1m)?
I've found AI too be pretty good at answering similar questions - so long as you form your prompts well.
> There's your problem: why continue to deny simple facts when you're shown them so clearly?
Who's in denial? You just completely skipped commenting on clearly visible twin external antenna sockets in gnss-compass version of the unit and didn't correct your wild assumption that smartphone sized brick is all that is needed for this gnss-compass to run.
And what's that stuff about size of all other sensors? Everybody knows those don't take any space at all and that every smartphone has them. My point was that having these additional sensors in a unit that boasts inertial navigational capabilities might mean that direction is also mostly innertially tracked since the last time the sensor moved or even mostly faked with magnetometer.
I can make better guess what (1m) is than AI. It tells us that this an accurancy is achieved when you put the two external antennas 1 meter apart.
Read Wikipedia page on GPS system and try to find any single feature unique to this system (or even just not common to all radio signals) that aids with determining orientation of the reciever (while stationary). You'll find none.
Who's in denial? You just completely skipped commenting on clearly visible twin external antenna sockets in gnss-compass version of the unit and didn't correct your wild assumption that smartphone sized brick is all that is needed for this gnss-compass to run.
And what's that stuff about size of all other sensors? Everybody knows those don't take any space at all and that every smartphone has them. My point was that having these additional sensors in a unit that boasts inertial navigational capabilities might mean that direction is also mostly innertially tracked since the last time the sensor moved or even mostly faked with magnetometer.
I can make better guess what (1m) is than AI. It tells us that this an accurancy is achieved when you put the two external antennas 1 meter apart.
Read Wikipedia page on GPS system and try to find any single feature unique to this system (or even just not common to all radio signals) that aids with determining orientation of the reciever (while stationary). You'll find none.
How do they remember to take it with them when they leave their home?
People with this level of dementia won't be able to live independently, and there will be someone else to check that.
Hopefully it computes a pedestrian safe path to your front door and points that way, not as the crow flies, otherwise it could walk you over a cliff or into a pond.
Id like to see a DIY variant! A couple of radios and homebase software. Could be a fun one to try an emulate.
If this device uses a similar technology as your phone, then I don't see it being reliable.