Smart Pedal

What's a typical spread when testing your MPG at each fill up? It's got to be a decent variable.
At least it is for me. Maybe ± 6%..... Maybe less....

 

But then your foot will also cause Less pedal travel the exact same amount when on this mythological 'bad road'.
Shouldn't that be a wash?
I can't visualize a road so rough that causes my right foot to move... but good theory....
What if the roads are smooth? What if it's always on CC when on the highway?

I'm sure you are seeing fantastic results using this gizmonic with your prius.
And you get a buckandahalf when chumps use your advertising link to purchase this 'amazing' device.

Say, if dampening the pedal's input to the ECU is so great,,,, why didn't toyota just add a few lines of code to do this?
Some like to say that MPG was always the most important thing in the Prius engineering.

 

That would be an empirical question ... you might have some personal opinion of whether your foot is moving the pedal, and an instrument comparing an accelerometer to the pedal position might reveal something different.

I largely share your skeptical instinct, but it's possible to overdo it and forget to distinguish one case from another. This thread is not about some product that dampens pedal input like a few lines of code could do; it's about a product that contains an accelerometer of its own, so it can do more than some simpleminded "dampening". The patents were linked in #27.

That still doesn't necessarily mean it's anything to write home about, but it's the sort of thing it could be interesting to see a fleet study of. The one long being anticipated in this thread seems to be out now (mentioned in post #56). I haven't had the time yet to read it myself, but from what's been quoted it sounds like the benefit was more modest than striking. The points from fuzzy1 about the small sample size were disappointing; I'd have expected those organizations to do something more rigorous.

 

My earlier comments were based only on the abstract. I had little time then and didn't locate the main paper. I have since located and briefly skimmed the main paper.

It briefly discusses other smart pedal designs from Nissan and Bosch, before moving on to this particular product.

The test "fleet" was 6 CalTrans pickup trucks: two F-150s, an F-250, an F-350, a Ram 1500, and a Ram 2500. A Bolt EV was initially included, but dropped due to data collection difficulties. The test period was approximately mid-August to mid-December 2022, two months without the device, then two months with it. Idle time and downhill grade miles (as determined by data loggers) were filtered out from the calculations. Total combined test fleet miles were 7475 miles without the device, and 10,313 miles with it, though I don't know if these totals are before or after the downslope filtering. Illustrations included mapped GPS tracks for each vehicle, and speed-binned MPG distributions for each vehicle.

That "average" savings of 6.29%, and payback period of 15.76 months, is not for the whole fleet, but for the single vehicle showing the best results, the F-250, which was driven 548.0 miles without the device, and 2793.4 with it. This vehicle had by far the fewest baseline miles without the device, and also the largest difference in miles driven without vs. with the device.

The other vehicles showed savings of 4.23, 3.02, 0.47, -0.52, and -1.72%:

"These small negative differences in average fuel economy are not expected to be reflective of the performance of the SmartPedalTM device since there is no mechanism for this effect, but rather reflective of uncontrolled nature of this study. [emphasis added] This study consisted of real-world operation and the contribution of unknown variables that could potentially impact fuel economy is unknown. Unknown factors that can impact fuel economy are discussed in Section 4.1, and include factors such as unknown payload, number of passengers, changes in driver, etc."

A lot more study is needed to determine if this device is actually beneficial, under what conditions, for which drivers:

"5.1 Recommendations and Future Research

Since the impact of the SmartPedalTM technology may be difficult to determine from a smaller sample size in an uncontrolled test under real-world conditions, additional testing would be needed to increase confidence in the effectiveness of the SmartPedalTM device and to characterize fuel savings with respect to other parameters such as vehicle speed. Two options for additional testing are a study with a much larger sample size performed under real-world conditions or a smaller study performed under more controlled conditions. In a controlled study, specific routes could be driven repeatedly, alternating between the baseline and the SmartPedalTM technology condition. The study could ensure that the driver, the number of passengers, vehicle payload, traffic conditions, road conditions, accessory usage, and other parameters that could potentially affect fuel economy, are as comparable as possible. Another area of study could also focus on driver behavior to determine if the SmartPedalTM device provides more value to drivers with certain behaviors."

 

Got some real good water front property in south Florida for sale cheap!

 

Checking and playing with a couple of their raw data files, these mile totals are very clearly before the negative slope miles are filtered out.

After negative slope filtering, this vehicle had 481.3 baseline miles without the device, and 2597.6 miles with the device.

I don't compute the same MPGs they display in Table 4.1. They show 12.88 MPG baseline, and 13.69 mpg with the device, a 6.29% gain. But with the 0-speed (idle) time and negative slope datapoints filtered out, per their stated method, I get 14.00 and [Edit: updated figure] 14.12 MPG, respectively, a 0.85% gain (before rounding) for the smartpedal.

With a bit of forensics, I can get their baseline figure of 12.88 MPG by filtering out just the negative slopes, and not the idle time, which accounts for 18550 seconds (5.15 hours) of the non-negative slope time and includes about 3 gallons of fuel burned while parked and idling. But I haven't yet found a way to get something close to their claimed 13.69 MPG from the SmartPedal dataset.

I hope I haven't made any serious spreadsheet errors in my calculations. The work is cumbersome enough that I haven't applied it to the other vehicles, and probably won't.

Will anyone else look at and play with these data files, to check my results or find other things?

===================================================
PS. For the curious, here are some plots illustrating the uncontrolled nature of these tests. For the F-250, this first plot is the relative amount of time spent at each speed. It is essentially a probability distribution function, X-axis is MPH, Y-axis is somewhat arbitrarily scaled (actually, % of the speed data falling in a 2 kph/wide bin, as the instrumentation clearly measured in integer kph, later converted to standard American units). From a histogram of the speed data logged each second, 11.4 hours for the Baseline set, 63.1 hours for the SmartPedal set:



Note that the Baseline spent more time at <10 mph and also >65 mph, both typically lower than average fuel economy, while the SmartPedal spent more time in the 30-50 mph range which would typically shows better fuel economy.

Here is the same data, re-plotted as a cumulative distribution function:

 

Fuzzy ! You are one data analyzing guy! Thanks for your report of this 'thing'.

Just simple seasonal differences can add up to more than ~6%.
Save your dollars on this gizmo, regardless of hearsay reports by forum members with a financial interest in this 'product'.

I can guarantee ~6% increase in MPG with no gizmo needed. Drive 5 mph slower everywhere you go! You're Welcome!

I'd say 80% of my driving is with my Adaptive Cruise Control engaged. This expensive little thing would just be along for the ride most of the time.

 

That sounds so much like "yup, that's what you'd have to do" that I confess I was thinking that was what they'd be doing the first time.

 

I hope that someone else will run their own analysis, to see if they get the same or similar result. I'm not error-free, and had to correct one number on this already, though the change was very small.

Yup. Even without changing drivers, which can have an even greater impact. These CalTrans vehicles likely had multiple drivers through the test period.

 

I guess when the funding grant is only $140693, one has to cut some corners somewhere.

 

Without hard data, user jibber jabber is commonly unable to identity uncontrolled conflating variables and very powerful placebo effects. Especially when there is some profit to be made.

Then there is also measured data that, upon close examination, turns out to be surprisingly soft.

 

I don't see why; companies that build a genuinely working product I want and profit by selling it to me, I've got no beef with at all. Most of the products I own I got that way. I didn't see fuzzy1 say anything different.

There's a useful skill set to have that gives you more choices than just "oh I believe that review" or "oh I don't believe that review."

You can also have the choice to, say, evaluate the review.

As fuzzy1 has so handily demonstrated.

 

Such as.....

"What's love got to do with it", T. Turner
Just a sweet emotion.

Show us hard data. A year's worth to deal with seasonal differences. Have you had this gizmodic that long?
If you want clicks on your links,,,so you can get in on that 'profit',,, you need to do more than talk about 'amazing products'.
How's that water injection going?

 

Wasn't it "a second-hand emotion"?

Or a "sweet old-fashioned notion"?