Motion-powered GPS tracker to install permanently inside a bicycle frame

[roybot]

TL;DR Summary

I could use some help designing a magnetic induction power source to convert kinetic energy to DC power, for use in keeping a GPS unit with GSM SIM card independently powered inside a bicycle frame.

Briefly, the problem; stolen bicycles are notoriously difficult to recover. Designing a GPS tracker to fit on one is challenging, because bicycles rarely have their own available power source. Whatever GPS unit is used must typically have its own battery as its sole source of power, which presents its own set of problems (e.g.; extracting/replacing, remembering to charge it). Stolen bikes also tend to get broken down quickly, so to be effective a tracker would need to be hidden inside the frame itself & out of sight.

I'm trying to design & assemble a GPS tracker meant to be installed inside the seat tube (not the seatpost) of a bicycle frame. If possible, I'd like to design it to use a magnet passing through coils, bouncing up & down across copper coils (like a 'shake' flashlight) to generate enough power over time to keep a small battery charged enough to enable a GPS tracker unit to - if needed - monitor its GSM unit and report its location.

Here's a link to a 'shake flashlight' project - the type of power source I'm talking about.

A GPS tracker unit is easy enough to find. There are tons of them on ebay/amazon. Most common are the types that have a GPS receiver unit with a GSM module featuring a SIM card slot. In all sorts of different ways, they use the GSM to receive signals from the maker's app, and the unit responds with logged GPS location(s). Usually there are settings for how often it logs on its own, or if it only should when asked, etc. Battery life will depend on those settings.

Some trackers promise standby battery life of 4 months or more - but nobody wants to have to dig out a GPS tracker from inside their frame every few months. Chances are they'd forget, and via Murphy's Law it'll get stolen as soon as they do.

I don't know yet how much power it'll take to keep the GSM cell phone unit in standby mode, but all of these units have a 'low power mode'. The GPS location logging can happen very occasionally.

I'll draw a diagram later, but basically I figure the thing will consist of;

1. The power source: a magnet inside a cylinder, springs on either end, and copper wire wrapped X times in the middle.
2. A 'diode bridge' to convert the AC power generated into DC.
3. A capacitor across the +/- after the diode bridge to 'smooth out' the power, some say (?).
4. [ some sort of component that takes weak and widely varying voltage and converts it to a steady 5, or 3.7, etc., VDC. ]
5. Some lithium battery power bank for the GPS unit, outputs 5V.
6. The GPS tracker unit, runs on 5V.

I figure this, intuitively, seems like a good use for a motion-powered battery, because unlike recent 'body motion charged' USB power banks (which have been a failure), it'll have one job, very little power will be expected of it, and it'll be embedded in a vehicle that experiences a ton of bumps & shakes.

If you're curious - yes, I am expecting that I'll need to connect the GPS and the GSM antennae to some point outside of the bike's frame - probably the bottle cage bolts.

 

[berkeman]

Welcome to PhysicsForums; this sounds like a fun project.

I don't think you will be able to get much energy out of a "shaker" type of energy harvesting, unless you mount the power generation unit to the unsprung parts of the bike (the lower part of the suspension that moves, which doesn't exist on rigid road bikes).

Instead, I'd recommend focusing on minimizing power consumption to extend battery life. You can do that with a sleep/wake circuit that consumes less than 1mA when asleep, and only wakes every minute or so to get position and transmit it. That's how we get battery powered IoT RF devices to last for a long time between battery changes/recharges.

You have a good idea to route the antenna(s) out to a disguised water bottle holder -- that's pretty clever and needed with metal frame bikes. You could probably put a disguised USB charger port in the water bottle holder as well, to make it easier to recharge the unit's battery. The signal transmitted every minute or so should include the battery charge voltage/state.

What is your cost target for this? What do you think the cost threshold for the bikes will be that this will be fitted to?

 

[anorlunda]

You may also be able to put a solar cell on the water bottle. You don't need much. You could cannibalize a solar garden light to get the solar cell, recharge circuit and battery holder all at once. Replace the light's LED with the GPS unit Check first on the required voltage for the GPS.

 

[roybot]

Thanks but no, for this to work it has to be absolutely out of view. Like I said, the lowest power consumption GPS trackers on the market last for 4 months max. But if someone can make a device like the 'Ampy Move' work at all, it should be possible to trickle-charge a battery enough to keep a GSM module on standby. It really shouldn't require much power to be stored at all. Extremely minimal power use is a given.

For instance, I've seen some GPS trackers work by powering off almost completely, then waking up their GSM and going onto the network for only for the first 5 minutes of every hour. The most efficient also only power on their GPS chipset to get a location fix when specifically asked to do so, and to do that you have to catch them between 2:00 and 2:05, etc.

The first thing most bike thieves do these days is chop a bike down to the components, often they'll even throw the frame away - it's usually the only part with a serial number. They're also very aware of GPS trackers. Anything that gives the slightest hint of containing electronics will be examined. Really the only option I see as viable is to place it deep in the seat tube, over the bottom bracket shell. Conveniently that lines it up with most frame's seat tube bottle cage mounts.

Anyways you can boil my project's ask down to 'how to make a linear motion based slow battery charger'. It's the taking of a very low and sporadic charge and storing it in a lithium battery.

... honestly, I might just buy an Ampy Move and hack it up.
I still want to figure out how to do this, though.

 

[anorlunda]

How about a secondary battery in parallel? You may be able to extend the lifetime on battery to a year or more.

 

[anorlunda]

If you're ready to sacrifice a few bike for experiment, try your concealment idea with the 4 month model GPS. Let them get stolen, then see if they transmit the whole 4 months, or are quickly destroyed by the thieves.

If I was a buyer and the seller was claiming that this GPS tracker fools thieves, I would demand proof for that claim. Technical details would not be convincing proof; only a real test with real thieves.

 

[hutchphd]

How about turning on the device only when motion of the frame is detected. The device would signal its location only if the owner's helmut was not In the vicinity (RFID?). This could mean very low quiescent current when bicycle parked and obviate electric shaker. I, too, believe the shaker WILL be unreliable and probably heavy.
If you already know your design (ie the magnetic pogo stick) just attach the full wave bridge to the battery...I doubt you need anything else

 

[Tom.G]

Summary:: I could use some help designing a magnetic induction power source to convert kinetic energy to DC power, for use in keeping a GPS unit with GSM SIM card independently powered inside a bicycle frame.

A capacitor across the +/- after the diode bridge to 'smooth out' the power, some say (?).

As @hutchphd mentioned, the capacitor is not needed. A rechargable battery acts as a VERY good, and large, capacitor!

 

[Borek]

This is public so you should be able to see even if you are not on fb:

(I am almost sure I have seen his post in English on some site but can't remember the details now).

In short, Tomasz makes GPS trackers that weight around 5g, use solar cells and supercapacitor (batteries fail at low temperatures up there), and flies them using small helium filled balloons. Some of them went around the world (literally). They don't use GSM but APRS, so it is a bit different, still, some ideas can be reusable.

No idea about technical details, but forces acting in the suspension seatpost can be large enough to produce quite a power.

 

[hutchphd]

There are piezoelectric devices that produce power from small amplitude oscillations which supply a few g of acceleration (they are little cantilevered piezo beam structures) but at low amplitude. These would be much better suited to this application...that's a great call. One company is called https://www.mouser.com/ProductDetail/Mide/S118-J1SS-1808YB?qs=9r4v7xj2LnlK6yfRfh%252BLlA%3D%3D&gclid=Cj0KCQiAh4j-BRCsARIsAGeV12CS_sFy6KvbV2sOffpmtTjRhIO-LktXyQEt4Ur6HNCXZFQ2HLYoGnQaAvosEALw_wcB

 

[Frodo]

As you park your bike at home, fix a contactless charger to the wall, and a receiving coil on the bike so as to charge the battery whenever the bike is at home. You could have a tiny socket on the bike and plug it in - in time, it would become an automatic action you wouldn't forget.

A problem with a hidden charger is you get no warning if it fails - you need some method of coping with this.

 

[mfb]

How about turning on the device only when motion of the frame is detected. The device would signal its location only if the owner's helmut was not In the vicinity (RFID?).

Then it needs to run the GPS and transmit continuously when it's in motion? That needs a lot of power.
I wouldn't trust the RFID detection (which is even more power).

Maybe send the position only after e.g. 5 minutes without motion. That way it's a single use per trip and you are sure to get the final position.

If I was a buyer and the seller was claiming that this GPS tracker fools thieves, I would demand proof for that claim. Technical details would not be convincing proof; only a real test with real thieves.

It's only security by obscurity anyway. It might work perfectly for a prototype, but if it gets more popular thieves will look for that.

 

[eq1]

I still want to figure out how to do this, though.

I don’t think the primary challenge is going to be technical. I think it will be economic.

After all, if money is no object just put an Apple Watch in the bike seat. It’s small, has gps and a cell radio, and great battery life with screen off. You can even write code for it.

If you do have a price target then that’s where the real engineering is going to need to happen and trade offs will need to be made.

If you’re just doing it to learn then what you want to learn will dictate the best solution.

 

[eq1]

Does attaching one of these to your bike solve the problem?

https://www.chewy.com/b/gps-activity-1899

 

[Frodo]

Before going too far check with your local police to see if they will assist you to recover it as they may deem it not worth their efforts to recover a bicycle.

 

[rbelli1]

If it is completely encased in the frame it will also not be able to receive the GPS signal.

BoB

 

[hutchphd]

Then it needs to run the GPS and transmit continuously when it's in motion? That needs a lot of power.
I wouldn't trust the RFID detection (which is even more power).

Not what I said. When it detects motion it would first search for helmut. If found: then go to sleep for maybe a half hour. If not found: emergency transmit position periodically until battery gone

 

[davenn]

Thanks but no, for this to work it has to be absolutely out of view.

well by definition it CANNOT be absolutely out of view.
You have 2 antennas that MUST be in the open else, 1) you won't receive satellites 2) you won't get a cellular signal

 

[Tom.G]

well by definition it CANNOT be absolutely out of view.
You have 2 antennas that MUST be in the open else, 1) you won't receive satellites 2) you won't get a cellular signal

With it completely out of sight in the frame, the metal frame acts as a shield blocking all radio signals... just like if YOU were in there, you would see nothing.

You could mount it in a tire, but then you have to balance the tire.
Or in the handle bars with the antennas in the hand grips, probably very poor antennas though.

 

[davenn]

With it completely out of sight in the frame, the metal frame acts as a shield blocking all radio signals... just like if YOU were in there, you would see nothing.

You could mount it in a tire, but then you have to balance the tire.
Or in the handle bars with the antennas in the hand grips, probably very poor antennas though.

Why on Earth are you quoting me ?
I don't need any convincing

Try quoting the OP Dave