Hi,

I've been doing some engineering analysis of key fobs:
http://hiwaay.net/~bzwilson/prius/pri_fob.html

Although North American key fobs work at 315 mHz +/- 2.5 mHz, there can be many different frequencies and data encoding methods. I'm using the Atmel, Blackbird development system to analyze the NHW11, 2001-03, and just starting on the NHW20, 2004-current, key fobs. Eventually I will be looking at a larger range of key fobs.

Questions?

Bob Wilson

 

Interesting, what do you hope to find out or discover? Just looking into because your curious?

 

My goal is to demonstrate a pedestrian safety key fob system for at risk pedestrians: (1) pre-school kids, (2) +65 age adults, (3) blind, and (4) mobility challenged pedestrians. To achieve this, I need to know how remote keyless entry systems work on my favorite vehicles. So I need to expand my network engineering skills with hands-on practice with:

• RKE band - 315 mHz +/- 2.5 mHz (FCC requirements)
• channel width - +/- 58 kHz (Atmel)
• modulation - ASK, FSK (Prius use ASK)
• data encoding - signal, NRZ, Manchester (Prius use Manchester)
• bit rate - 615 and 713 bits/sec (Prius bit rates)
• message - ~36 to ~48 bytes (Prius message sizes)
• power budget - 10 ma. @ 3.3 V * message_length (Atmel)
• processing requirements (under investigation)

Having proposed a pedestrian safety, key fob, I want to build and test systems for head-to-head, field evaluation. This can lead to accurate traffic models, an optimum solution and effective vehicle safety regulations.

The blind have a bill, HR 5734, that proposed to put noise makers on our hybrids without alerting the driver about at these at-risk pedestrians. The sound might notify a blind pedestrian but it is also likely to be lost in the traffic. The proposed 'noise makers' add costs to our Prius with no proven safety improvement. But I look at it from a different point of view.

We can expand the utility of the existing key fob receivers by adding new software to eliminate hardware cost and expand coverage to a larger pool of at-risk pedestrians. This can demonstratably increase safety and may be enough to get an insurance policy discount for our Prius.

Bob Wilson

 

So if I get what you're saying, if the adopted design parameters go toward modifying existing key fobs through software augmentation, an at-risk pedestrian would then be able to buy one of the generic key fobs that would already be in distribution to car owners, carry it in their pocket, and the proximity of a hybrid vehicle would set off an alarm noise from the vehicle? Sounds like it would work if the alarm was the horn. If it uses the normal positive feedback beeper of the key fob, the pedestrian may not notice it.

Of course, it's really up to the driver to avoid pedestrians anyway. All this does is give participant pedestrians an audible to replace engine noise in their hazard recognition.

Seems like over kill to me, but hey! If the insurance companies think it is worth a discount who am I to argue.

 

Exactly, the horn but a short 'bleep'. It is called "pulse modulation" and even with the horn relay, this should work quite nicely. But I see the special key fobs as being attached to shoes of pre-school kids, senior citizens and other at-risk pedestrians.
Can't you hear your own horn? I think the driver, pedestrian and any by-standers will get an audio alert. For good measure, flash the lights too.
Better still, it scales well. All key fob equipped vehicles can use the same system, ~80% of all vehicles. What this means is even the remaining 20% get some notice of the at-risk pedestrians.

Bob Wilson

 

Wow. That is an incredible idea. I really hope this could be implemented some day. I think this would also be useful for other luxury ICE vehicles that can run just as quietly as a Prius in stealth mode. My coworker's new Lexus is silent when not accelerating.

 

This is a great idea. I would like to offer a different spin on it.

What is the car emitted an RF signal that could be picked up by a fob on a person?

Some advantages i see:
1. The fob could do different things for different needs (vibrate for deaf persons)
2. Battery life would be longer of the fob.
3. Would not create "extra noise"
4. The transmitter could be attached to other vehicles (like bicycles or scooters)

 

How are you going to be able to receive the fob's signal? I've noticed several times the range of my key fob is only a few feet from my prius. is this a weakness of the antenna or receiving system or did they do it on purpose. Anyway what do you consider a acceptable distance to alert a pedestrian?

 

You are right but this requires bi-directional communications. Most cars today only have the receiver. Only a few cars have bi-directional communications and the only deployed system I know of in North America uses a different frequency band for the car-to-fob channel.

The NHW20 uses a 134 mHz signal from the door handles to query for a near-by keyfob. The keyfob still transmits in the 315 mHz range. But this primitive system is just a fraction of the installed base.

The keyfobs I've looked at have terrible wireless efficiency. Their bit rates are abysmal and there is no 'protocol' to avoid collisions. However, these receivers are installed in 80% of all North American vehicles. So this is how I would like to approach the problem.

PHASE I - SAME MAKE AND MODEL PEDESTRIAN KEYFOB

The first version is required for all vehicles and requires a pedestrian keyfob with these features:

1. common modulation - everyone uses ASK (on/off)
2. manchester encoding - self-clocking and zero average signal level
3. fixed bit rate - 38.4 kbaud
4. fixed packet format - sync, identifier, certificate, Rx freq, check
5. defined packet TX pattern - redundant, semi-random packet burst

All of these elements can be implemented with software within the existing keyfob systems. The NHTSA and other, open source authors, would publish code to implement the above protocol. Every manufacturer would have a short response time, ~3 years, to implement this first version and offer a pedestrian keyfob with each vehicle.

PHASE II - UNIVERSAL PEDESTRIAN KEYFOB

Working with the FCC, one narrow band in the 315 mHz range would defined as the beaconing frequency. Delayed ~6 years, all manufactures would be required to implement:

1. vehicle monitor the <tbd> beaconing frequency (manditory)
2. vehicle TX beacon on that frequency using a carrier sense, collision avoidance scheme (optional)
3. pedestrian keyfob with this capability with each car (manditory)

This verson requires a frequency agile receiver in each car, often a minor circuit tweak that can needs to be software controlled. The optional TX capability requires a significant change to the vehicle electronics that now must also be a transmitter. This is not a trivial problem, which is why a longer lead time is needed. However, keyfobs are not the only and in the current version, not the best approach to adaptive, wireless networks.

PHASE II - ALTERNATIVE WIFI OR CELLPHONE BASED

I would also like to give the existing, wireless industry a chance to propose a keyfob replacement based upon:

• cellphone - this provides a much more sophisticated 'keyfob', a cell phone and the ability to treat cars as an adaptive, cellphone network. Think of it as open source, On-Star, and this is the technology I would prefer to see.
• WiFi based - this technology is racing ahead and has always been data centric but a WiFi based keyfob would be of limited utility since it could not place a 911 call. Still, it solves the protocol and data problems very nicely. This is the European approach 'Cooperative Vehicle-Infrastructure Systems.'

I see Phase I as saying, "either come up with a better solution or Phase II is coming." This would let the true wireless experts, the cellphone and WiFi community, get access to the automotive base. Finally we would have a wireless data network capable of handling both pedestrian and vehicle safety.

Bob Wilson

 

Testing in a parking lot showed my NHW11 responding to my keyfob at +100 meters.

Bob Wilson