Right to Repair

Hey CD,

I’ve been thinking a lot about the Right to Repair movement, and in general designing things to be repairable after some of the catastrophes at our first competition last year. I personally am a big advocate, and really believe in the idea that anyone should be able to repair their own possessions. But at the same time, I like how thin and light modern electronics have become. What do you think?


I’ll give a a personal perspective on it: Many products are engineered to perform a very specific task, and certified to do it safely and securely.

Opening up the possibility of repair also implies the ability to modify behavior, in potentially unpredictable ways.

In a large mechanical system, it’s often very obvious to visually inspect for modifications, and observe that behavior is correct. In software and electronic systems, it’s not always as obvious. I’m painting with broad strokes here, but I think it’s a generally true set of statements.

For that reason, if I’m tasked to produce a system that is compliant with safety or security regulations, part of my design criteria is the ability to prove that my functionality is the one running on the device, not an altered form of the functionality. This often implies a reduction in serviceability.

Do I believe people should be able to diagnose and fix problems in devices? Yup, for sure! Don’t throw away something that’s 99% working! That’s wasteful.

Do I believe anyone should have the ability to tinker with their car’s ECU? Definitely not. It’s dangerous, and frankly, I don’t trust people.

That being said, I’d definitely like to have more transparency than what we currently have. Open source software and hardware are some of my favorite things. However, there always has to be a balance, to ensure uninformed people don’t hurt themselves.


Indeed. Right to Repair should be limited to those who know what they can do and what is safe. Many repairs require one to be careful and understand the system. Not everyone can do that.

1 Like

a few different points here.

  1. There is a difference between designing something to be repaired and designing something to be repaired by anyone. You used electronics as an example. An example of a repair-ability feature in electronics would be not placing tons of 0201 passives next to a BGA. This would make the board far more repairable by a person with the tools and experience as the risk of messing up the solder on the 0201 parts while removing the BGA and reinstalling it is minimized. A lay person obviously would not be able to repair this regardless, but it is still designed to be repairable by a technician.

I believe parts should be designed to be repaired by technician, as the lifecycle of a part should not involve binning it after it fails because it would be too costly to fix.

  1. This echoes @gerthworm a bit. Engineers design things in very specific ways, and they are assembled with correct processes that were also engineered to ensure that the product is a conforming product under a formalized quality system. Thus taking it apart violates all of that quality insurance leaving the consumer to risk of product or personal damage. In America litigation against companies is a sport, and with certain legal precedents that may be set soon (Sandyhook parents v Remington) The ability to litigate a company due to injury caused by misuse of that companies product could be a real thing, and thus force companies to be even more cautious of allowing their products to be modified.

I wish people would take more personal responsibility in the matter, but I fail to see any restraint in terms of the current lawsuits brought up against companies for things like not labeling a hot item.

I agree with you that anyone should be able to repair their own possessions, but I do not believe it is a companies responsibility to make it repairable to the lay person, and I fear consequences brought on by the increased liability companies may face for allowing consumers the ability to repair their products.


It’s one thing to repair a mechanism that your team designed that is damaged or doesn’t function as planned.

It’s something completely different to try repair a battery, PDP, rio, pressure relief valve, motor, or most other COTS parts.

What exactly are you trying to repair?

I think everyone deserves to at least have a shot at repairing their own stuff, especially since warranties only last for so long. Design decisions that make it harder to do so are usually either very deliberate for any number of reasons (aesthetic, packaging, appeasing corporate, “less connectors”) or the result of bad engineering.

If I wasn’t able to take apart toys I had lost interest in when I was younger, I don’t know if I would be posting this today.

counter hypothetical: if you took apart a toy without knowing of a certain spring, you may be posting with one eye today.

I’d say that for some things, such as laptops, people should at least have the right to upgrade them without buying a completely new thing. For example, replacing an SSD or upgrading RAM. There’s a reason I like my 2007 17" MacBook Pro.
However, it wouldn’t be smart to let everyone try to replace their phone batteries. coughs Samsung Note 7 explosions, anyone?
I’m all for open source stuff, but not necessarily everyone should be able to repair some ridiculously complex extremely fragile thing.

Or you replace your phone screen (not covered under warranty) and then drop it into water. The company should not replace it as you probably did not apply the seal correctly. There are issues that are your fault. Or what happens if you break for motherboard in the process and then try to pin the issue on the company. They have to deal with you.

When I worked for a small DC motor company the owner was telling me a story one day on our carpool down to the production plant. That they had a customer that had sent back several motors because they were under performing. On return they noticed that the motor had been opened up by the customer.

It turns out the customer was dissembling the motor and throwing out the front end bell just to assemble it directly onto a gear head. Through this process they messed up the seating of the brushes, and it was causing some power loss.

TLDR: Taking something apart without proper knowledge, or tools can lead to several product or personal damages. ESD shock would probably be the largest for those trying to repair electronics.

1 Like

This is where I do get quite torn. There’s two sides to this always, as illustrated by a recent incident involving John Deere and right to repair. As an engineer in this space, I want to be really careful who gets to tinker with the innards of my software. At the same time, as a consumer, I get frustrated by being forced to go to the official company-certified repair technician to do a repair I know I could fix myself.

Honestly, I don’t mind how the automotive industry generally does it at a high level. Lots of the basic routine things can be done at home. Some heavier replacements are better suited for a general technician with better tools. The weirdest, most-detailed problem solving gets done by the car-specific service team at the dealer. The consumer generally has the freedom to choose where they attempt the service at, and service techs generally can dictate “what am I willing to try to fix, what am I not”.

EDIT: To those reading the article, it’s important to remember that when they talk about “Car mechanics reprogramming controllers”, they’re referring to settings changes, not writing and uploading new source code. Think “change wifi network settings”, not “Write new wifi driver”.

EDIT2: I should also say - the article is heavily biased against John Deere, and doesn’t dig into the nuances of what customers in that industry expect. It probably wasn’t the best example, but at least it’s a sample point of the level of public discourse that currently exists.


Ah yes, the “I trust myself to fix this properly, but don’t trust my neighbor to not roll coal” approach. :grinning:

On a more serious note, most product design decisions are made from a desire to achieve a certain set of features at a certain price point, while meeting all legal obligations/regulations. This value engineering approach generally wasn’t some evil mandate[1] from above, but rather the resulting organic side effect of how people actually open up their wallets and spend money.

Essentially, the average consumer[2] of a good or service generally cares more about the sticker price of an item, rather than the TCO over the lifespan of the product. Making products more repairable almost always results in higher initial product costs[3] for consumer electronics type products (cars or appliances are generally not too difficult to repair).

While repairable products would generally have a lower long-term total ownership cost (and lower environmental footprint) to the consumer, most consumers generally don’t respond to this and other “invisible” negative externalities. For example, a few percent of the population(who might self-describe themselves as “woke”) bought hybrid or electric cars to help the environment, whereas a lot more are buying electric cars because the sticker price has dropped, driving performance is great, range has reached acceptable levels, and charging costs are a lot lower than a gasoline fill up.

The only way to correct a negative externality in a free market is a government regulation that forces the costs of that externality to be reflected in the sticker price paid. Then consumers would organically choose the “better” option as the sticker price would be cheaper. One perfect example of this would be a carbon tax and return dividend proposal to solve climate change.

But as already noted elsewhere in this post, new government action can easily have a good intention, but could miss the mark and cause more problems than it attempts to solve. So it’s a tight balancing act that would require a self-correcting system to tweak and iterate towards a genuine better solution.

And as others have noted, as we achieve higher and higher levels of automation, safety becomes an increasingly paramount issue. Should people have a right to reflash their own firmware into a Level 5 self-driving vehicle? If this modified vehicle hits and kills someone, how to you identify who is at fault? Should modified L5 autonomous firmware (that hasn’t been extensively tested and verified as safe) be treated the same as a DUI (e.g. you are enabling control of a vehicle that can easily maim and kill others without being fully in control of the situation)?

On a personal note, I’ve repaired plenty of my own items if it makes sense (e.g. I’m not resoldering a BGA component on a PCB, but I have replaced batteries in devices), love watching AvE on YouTube (especially the BOLTRs on product design), and have generally shifted towards a more Marie Kondo / pro-sustainability approach of only buying a limited amount of high quality stuff that will last a long time. While I don’t know what the right answer for the world at large is (does anyone?), at least I have control over my personal decisions to be more individually sustainable.

TL;DR: This is a complicated issue that almost certainly cannot be resolved through an unfettered free market alone.

[1] That’s not to say these have never existed. One clear example was the Phoebus cartel, another arguable one was the GM streetcar conspiracy, but this could also be argued as hastening an action that would have occurred anyway. E.g. the only cities that kept their streetcar systems operational to the present (Boston, Philadelphia, Pittsburgh, Newark, San Francisco) all survived because these routes operated on mostly dedicated rights of way with subway tunnels in the downtown core. Almost all of the shared right-of-way streetcar lines, even in these cities, were eventually converted to bus. Another contributing factor was over government regulation, as many privately-owned streetcar companies were required to perform all maintenance on the roads they operated on, even for the automobile lanes, while bus companies did not have to pay for any road maintenance fees to operate.

[2] Most Americans don’t even have $1000 in savings to cover an emergency expense, and thus care a lot more about the sticker price of a consumer good than ongoing ownership costs.

[3] Existing government actions (like property taxes on warehouse inventory) heavily incentivize businesses to make everything just-in-time, with little actual inventory. It’s expensive to make a few widgets at a time, but pretty cheap to make thousands at a time, and even cheaper to make millions at a time, so this incentivizes companies towards a few high volume SKUs. Stocking a few pieces of every subcomponent to be repaired for individual sale is almost always unprofitable - which is why sometimes spare parts (such as for power tools) that do exist might cost 60% of the cost of a new complete widget altogether.


Super awesome response!

:smiley: When I say “Don’t trust”, this is really what I’m getting at (though I have an irrational hatred for intentional coal rollers). This lack of trust is strongly motivated by my day job. I’m far from an expert in the field of ECU repair rights, but I do believe I have a unique inside perspective.

1 Like

All that ^^ is good.
With skilled labor charged at $125 an hour*, a 30 minute repair on a $50 item doesn’t make economic sense. My labor is free. I guarantee I can get it apart…

*My car dealer charges ‘only’ $125 an hour.


Sounds like fun!

1 Like

It is difficult to draw the line. My Nissan kept throwing codes implying one of the catalytic converters and one of the air fuel sensors was bad even though both were recently replaced with new components. After replacing the cat and sensors twice, I eventually found a very small air leak between two of the bolted flanges. On some vehicles, it is possible to extract the sensor data from the ECU to quickly and easily diagnose this type of situation. For Nissans, extracting this data requires one of their proprietary interfaces that no regular consumer could justify purchasing.

Two thoughts:

First, there’s an art to choosing the exact text message that goes along with a particular diagnostic code to drive the proper behavior.

Additionally, In your particular case, I think both sensors were subject to in-range failures - conditions that electrically appear as “faulted” are achievable during reasonable operation. Maybe not “normal” operation, but a small hole in a pipe is far from an unreasonable failure.

Of course, no offense meant to any Nissan engineers here. Hindsight is always 20-20. I’ve had my fair share of misses too.

Tying this back to the discussion at hand: Getting the system design right is a key portion to preventing “right to repair” issues from becoming, well, issues. If you make a product that rarely needs repair, this is suddenly a much less important conversation.

Now you’ve got me interested.

But not interested enough to join linkedin.

I’ve been “retired” long enough that I haven’t really paid much attention to “right to repair”, since I was fixing cars, and folks in the independent repair business were pushing to get more info about how to deal with the pleasures of working on the then-new OBDII stuff.

Being old, I’ve kind of given up on ever being able to improve lousy built in software, on anything. Or even lousy “updates” on open source software, such as when my phone becomes less user friendly with the latest feature/security fix update. And I’ve become much better at throwing away stuff that should be good for a lot more use, but isn’t because of either intentional or unintentional unrepairability.

I love Art’s post above…