Native vehicle-bus decoder


#1

Hey everyone! I’m part of a tiny company working to make a better vehicle decoder. We’re curious to see if there’s any interest in what we’re making, and what our process is. Our intent for developing our platform is to provide data that is inaccessible by OBD2-only scantools. Our method is passive. This allows for us to decode far larger volumes of data without taxing the vehicle by injecting traffic onto an already crowded bus.

The OBD2-only protocol is commonly used to extract data, but cannot decode things like door position, state of lighting, state of HVAC, mass airflow, shifter, current/desired gear, steering wheel angle, turn signals, brake switch and more.

I’d like to spark a discussion about the applicable telematic applications that people might want to see, how these parameters would be best displayed, and anything else that might come to mind.

Currently, we have plans for performance enhancement, fuel efficiency, safety, alerts and reminders, data logging, recreation, car customization, and so much more. We will be adding compatibility for third-party applications, and potentially integrating third-party hardware. In the distant future, we hope to become a Vehicle to Vehicle communications platform.

Edit:
This is a telematics tool designed to be used with recordings as well as live data. Our focus is to decode at very high data rates without injecting requests onto the vehicle like OBD2 does. We can record and analyze persistent data over time e.g. tire pressure to determine if you might have a slow leak.

For example, on our test vehicle (a 2010 Infiniti G37), the TPMS (tire pressure measurement system) reads and stores tire pressures, but they cannot be viewed on the vehicle’s multifunction displays. Our system can read and display data on an external device (e.g. tablet, cell phone) to increase the functionality of the car.

Thank you!


#2

I don’t need a machine to check those things. I can just look at a door to determine if it is open or closed. The same goes for turn signals .


#3

You’re absolutely right, but there may be times that you’re not in the driver’s seat, and would like to know how well someone else is driving your car. If your teenager isn’t signaling while merging, or if a trusted friend you lend your car to abuses that trust to street-race, it’s hard to tell exactly what they’re doing with a regular scantool.


#4

Please don’t take this the wrong way . . .

OBD2 has NOTHING to do with lighting, doors, and most of those other things you mentioned

The solution is simple . . . buy a professional level scan tool

And I agree with Volvo . . . you can check lighting by turning on the lights, getting out of the car and checking if they’re working

My mom’s civic says exactly which door is open. Don’t need a tool for that

Why try to fix a problem that does not exist . . . ?!


#5

You don’t need to know if a teenager is doing something wrong in a vehicle. You can just count on it . And you don’t loan your vehicle to someone that would abuse your vehicle in the first place.
Also the insurance companies already have tracking devices for erratic driving.


#6

Shawn:
I think I know what product you’re building.

I encourage you to figure out what data is valuable (and hence worth someone paying $$ for) vs data that just creates clutter.


#7

You’re correct @JoeMario! I didn’t feel it would be appropriate to link, I have zero intention of begging or spamming people to throw money at us. We’re trying to improve our understanding of what the market currently lacks, and how we can best focus on developing new tools and applications. We’re currently exploring Dashware to display data graphically, and we know that it’s not easy to decipher our current data ouput at a glance.

@db4690, and @VOLVO_V70

I really appreciate the input from both of you!

I’d like to add something else to consider: A universal platform does not currently exist for automobiles. If you want to sync up additional devices and manage them from a single hub- you’re out of luck. There’s a large number of applications that could be very interesting to use with a vehicle, but are difficult to implement for a casual user. If you care about maximizing performance, there’s data that’s simply unavailable with even professional scan-tools.

Edit: I was incorrect, and professional scantools may indeed have specific proprietary data available, but are generally not universal, and cannot be used across any make/model.

Edit2: Additionally, please note that the price tag on professional scan-tools that include proprietary data (manufacturer owned) may cost thousands of dollars. Our solution is significantly more affordable.


#8

Fuel trim would be a nice addition.


#9

@Barkydog Long term and short term fuel trim are already accessible through OBD2, so they’re readily available within our prototype! We didn’t have a lot of space to work with when creating our initial video overlay, but we can access any parameter that the OBD2-only protocol supports. Thank you for the suggestion!


#10

If your friend is so disrespectful that he would street race YOUR car without your knowledge or permission . . .
he’s not exactly a high-quality friend

And if you’re the type that WANTS to spy on your own friends, what does that say about you . . . ?!


#11

The most frequent complaint from get from car owners I’d guess is “the engine won’t crank”. The most common reasons are (in no particular order)

  • automatic transmission safety switch faulty
  • clutch safety switch faulty
  • battery faulty
  • battery connections corroded
  • starter motor (or starter solenoid contacts) faulty

To the extent your device could quickly diagnose that problem, it would be very useful. Even better if electronics could be installed on the vehicle and predict when a no-crank was about to happen. Owners really don’t like to come back to their car and find they are stranded due to a no-crank.


#12

@db4690 I think everyone would agree with you! I’d like to give a similar example to illustrate our intent. If you were running a small business which included a small fleet (even a single vehicle for pizza delivery) would you like to know if your employee was driving unsafely? With a dashcam in the vehicle, it can provide some details regarding an accident- but is limited on the angle of the camera, and requires the user-performance review to analyze a video. Videos are very data intensive, so storage can become a major issue. With a data-logger (potentially with access to a GPS, dashcam, and other assorted sensors) a performance review could be as simple as looking for warning alerts about lane changes without signaling, driving over the speed limit, and any other unsafe driving usage customizable to the user.

@George_San_Jose1 You bring up an excellent use-case for how a diagnostic application could be useful. At our current prototype phase, we have too limited sets of test-drives to accurately discern when your listed specific warning codes may occur. We will further investigate this line of inquiry, please note that our current primary focus for our particular product is telematic, and not diagnostic- though certainly, a viable and desirable goal to pursue!


#13

So all the info would be available to your tablet, phone etc.? I think by the time you put in enough intrusion security it would no longer be readily affordable. Besides doesn’t OnStar and others already do this?


#14

If you go to GITHUB - you’ll find an app already written that probably does 95% of what you want. It’s for Linux written in C.


#15

How is this thing going to hook into the car to provide OBD data as well as the other stuff like doors and lights?

Is it a permanent installation?

While I agree in principle that I can see what door is open, sometimes that’s not the case, as when a door is almost completely shut but didn’t catch that secondary latch. I can see this as being useful in the few cars that don’t already have a door ajar warning from the factory.

That aside, I take your point that it would be nice to know what’s going on in all of the systems of my car. For instance, I know the navigation system knows how fast I’m going, but it won’t display it, which would be a very handy rough speedo calibration check.

The big question is how much this thing is going to cost. If it costs as much as a professional scan tool then you won’t sell many because as a non-professional I’m not going to spend that kind of money when I can spend $20 and get Torque and an OBDII adapter and get 90% of what I want out of a scan tool. And I can’t imagine professionals giving a rip about whether the door is open, and they’ll keep checking lights the way they always have and stick with their pro-grade scan tool.

What I’d really like out of a consumer level scan tool is one that natively recognizes all of the unique vehicle codes. Torque makes the attempt - you can install extensions to recognize certain codes from certain manufacturers, but it doesn’t go far enough.

A more easy interpretation of the codes (i.e., I never remember which bank is bank 1 without popping the hood and removing the engine cover to see where the cylinders are, which isn’t convenient when trying to diagnose a friend’s car that I’m not familiar with) would be lovely as well. Maybe instead of just “bank 1 sensor 3,” it shows an actual top-down picture of the engine with a red dot to note the location.


#16

George, please be realistic

No single device is going to tell you if all of those items are corroded or faulty

You’re going to have to pop the hood, look at the connections, use a multimeter, battery tester, common sense, and have a decent knowledge of how automotive electronics work

Better to avoid this situation altogether

Replace batteries proactively every 5 years, as I do at my job

You’re late to the party . . . to be blunt

Such devices already exist. They track the vehicle’s location, the vehicle speed, how much fuel was put in the fuel tank, etc.

They’re used by professional fleets, usually very large fleets, to be exact

Why waste manpower trying to develop something that already exists, and does EVERYTHING you mentioned, and several things you didn’t, to really make a point


#17

If this guy can do it a lot cheaper then he’s got a market. Otherwise, you’re entirely correct. Not that I like the idea of monitoring my employees that strictly. If I were running that fleet business, I’d hire the best people I could find, pay them a fair wage, and expect good behavior out of them. I’m not going to monitor them every second of every day because that’s a great way to make good workers feel like crap and quit.


#18

Many of the larger trucking companies . . . maybe all of them by now, come to think of it . . . use devices so sophisticated, pretty much every single drivetrain parameter is monitored and recorded.

If the guy makes a bad gear change, if he should be in a higher gear, vehicle speed, everything imaginable

And supposedly the guy will be told he’s driving in a manner which will shorten the life of the equipment and/or in a way which is less than 100% efficient, thus wasting fuel. Apparently, it all boils down to money.

Our fleet uses devices which aren’t quite that sophisticated, but they do record quite a few things. And some of the employees have been lectured for exceeding the speed limit or being somewhere they shouldn’t be. One guy was criticized for being one mile off of his assigned route. When he explained he was getting lunch, it apparently wasn’t a valid reason . . .

Some of the operators are assigned to areas that are . . . shall we say immoral . . . and the fleet doesn’t want any of their guys caught dead to rights doing something that is shady, if not downright indecent or illegal


#19

Our project lead has developed J1708/J1587, J1939 and OBD2 decoders for several products that are currently used in the telematics industry. Our decoder has been developed to address the shortcomings that we have seen in the field. In the past, we have needed to hardwire connections to turn signals, brake switches and ignition systems because these protocols did not have the data available. OBD2 was never designed for telematics use. You cannot pull data from an OBD2 system reliably faster than about 100 parameters / second before you start interfering with the operation of the car. For example, we tested a 2009 Saturn Aura that started to throw DTCs and lit up the dashboard like a Christmas tree at 75 requests per second (15 millisecond intervals). On the other hand, running on a cheap Raspberry Pi, we can pull 6000 parameters per second passively, with no possibility of interference in real time.

This is not to say that OBD2 is useless, far from it. It is just another tool in the toolbox. We can seamlessly integrate both. Vehicle tech is advancing quickly and everyone puts in proprietary methods of dealing with the sensor fusion problem. The manufacturers put the support effort into their protocols to add new functionality, not OBD2 which is usually an afterthought.

@shadowfax

We use off-the-shelf parts with a Raspberry Pi and Socketcan capable CAN-adapters. It is up to the user if they want a permanent installation. We pull our dashcam after every drive, and do the same with the data recorder. We’re currently evaluating some industrial strength platforms for permanent installation, but they are much more expensive.

We have worked for several companies that provide telematics and safety applications for heavy vehicles. The trend is for greater integration of recorder systems with external sensors (cameras, radar, sonar, etc). The big motivation is risk-aversion and after accident forensic analysis. Most of the systems have at least 2 cameras, one forward for lane and vehicle detection and another one looking back into the cabin. The cabin camera has the potential for use for things like driver awareness detection. And yes, it does bug the crap out of the drivers. Many of them hate it until the camera exonerates them from the responsibility of a collision. It is all about reducing risk to the fleet operator and the drivers.


#20

you remove (pull) your electronics after every drive, which could be many times a day? why would you do that, seems expensive and intrusive.