Category Archives: Uncategorized

RE: The “DIY Self-Driving Car.”

According to a semi-recent article from the consulting firm IHS Markit, the average length of car ownership in the United States for new cars and light trucks is right around six-and-a-half years, while the average age of all vehicles on the nation’s highways sits at around eleven-and-a-half years.  On its own, this might not seem like an important piece of information unless –– of course –– you’re an automotive manufacturer hoping to keep the conveyor belt of new car sales in constant motion.  But this is an incredibly important piece of information to keep in mind as advances in self-driving car technology make rapid improvements.  In fact, let’s use this information as the springboard for a quick hypothetical question:

What if it were less expensive to retrofit an already existing/already purchased car with autonomous technology, rather than buying one that’s brand new and with built in self-driving capabilities?

In other words: what if you could just go out and purchase an aftermarket DIY self-driving car kit that you (or, more likely, a certified mechanic) bolted on to a car you currently own?

Well….  You can.

Kinda’ sorta.

Though the technology is still in its infancy (because all self-driving technology is in its infancy), several companies have tossed their hats into the DIY autonomous vehicle ring, and are making “bolt on” hardware and downloadable software available at incredibly reasonable prices.  At the time this post is being written (May of 2017) the following companies/start-ups are offering –– or are working on –– some form of aftermarket autonomous technology, designed to be incorporated into cars that would normally require human guidance, steering, and attention at all times:


There’s nothing intuitive about the AutonomouStuff website, though a deep dive into the About page and, especially, their “Products” page reveals that the company is selling both aftermarket parts and hardware for building, testing, and developing autonomous vehicles.  Think of these guys as kind of a “Radio Shack for DIY self-driving car enthusiasts,” with the added benefit of white papers, research, and software downloads.  Again, the site’s a  bit confusing, though here would be a good place to start.

George Hotz made his first big splash in the media as a teenage tech phenom, gaining notice for a set of room scanning robots he designed, built, and entered into the Intel International Science and Engineering Fair competition (and becoming a finalist in the competition).  Later, he became know as “that guy” who hacked his way into Apple’s iPhone and Sony’s PS3 gaming platform (and then end up being sued by Sony).

A jawdroppingly-intelligent guy, with little need or respect for the rules and expectations of society, Mr. Hotz stunned the world technology back in 2016, when he announced his self-driving platform –– a “DIY semi-autonomous car kit” that he would make available to the public for $999.  Unfortunately, the rules and expectations of society caught up with George in the form of a letter from the National Highway and Traffic Safety Administration (NHTSA), asking that the Comma One platform (the car kit being offered by undergo rigorous safety testing before being offered to the public for sale.

Again, rules and regulations aren’t really George Hotz’s primary concern, and so –– rather than submitting to the government’s requests –– the Comma One software is now being given away for free (now as openpilot), with the hardware available as an open source list of parts and schematic boards.

Created by a group of ex-Stanford researchers, is a Silicon Valley startup using machine learning to teach self-driving cars how they should interact with passengers and pedestrians.  At the moment (May of 2017), there doesn’t appear to be any specific product offering from the firm.


Don’t want to deal with parts, schematics, and a soldering iron?  The folks over at Neodriven have taken the open sources parts lists and schematic boards from and are building semi-autonomous hardware kits based up the plans.  Available for purchase for the either $999 or one bitcoin.

Perrone Robotics

The development  team behind Perrone Robotics has had –– relatively speaking –– a long history with autonomous vehicles, and was an early participant in the DARPA challenge, entering their first vehicle into the government sponsored rally back in 2005.  They currently offer a software product, MAX, which can be configured for passenger cars and smaller robotic vehicles and devices (i.e.: delivery bots, golf carts, warehouse inventory transport, etc…).  One especially interesting aspect of Perrone Robotics and their management team is they’ve recently brought Atari-founder Nolan Bushnell onboard as an advisor.


The Oregon-based company, PolySync, offers the basic hardware and software for building and testing a self-driving “by-wire technology system,” designed for use on 2014 (and later model) Kia Souls.  The basic PolySync hardware set-up can be purchased directly from the company for less than $1000, with full instructions on installation and software available via GitHub.

Obviously, when looking through the list of companies above, it quickly becomes apparent that most of the hardware kits and software available for building a DIY autonomous vehicle are in their development and/or experimental stage, with little commercial release.  This makes sense, of course, because self-driving cars are in their earliest stages of development.  Do realize, though, that autonomous vehicle technology is an exponential technology: the early advances start off slow and then –– BOOM! –– it’s like a rocket taking off.

Now, taking a moment to relate back somewhat to the purpose of this website, what legal issues might be associated with a self-driving car car assembled at home or using some aftermarket kit?  Let’s say you were to purchase an autonomous car kit of some sort, have it professionally installed by a certified mechanic onto a car you already own, and then –– some time after the installation –– the car gets into an accident?

Where might the responsibility lie?

Would there be some onus of responsibility on the mechanic/installer, even if that person were certified is some way with the technology?  And what about the original manufacturer of the LIDAR and camera vision systems?  Or what about the software that navigates the car?

These questions need to be asked now –– not with any real answer in mind –– but with the knowledge that these questions will be asked over and over during the course of the next few years as more autonomous vehicles make their ways onto the roads and highways.   And the overaching theme of responsibility will come up continuosly when discussing semi-autonomous vehicles, fully autonomous cars, and the various forms of aftermarket autonomy products.


Decisions, Deaths, and Self-Driving Cars…

At some point in the very near future, an accident will take place with a self-driving car, where –– with clear purpose, though with no intended malice –– the on-board processors, software, and networked guidance of the vehicle will decide to kill someone.

Or kill an entire carload of people.

Or wipe out a few pedestrians making their way across a crosswalk.

And it will make this decision based upon logic, circumstances, the laws of physics, and an incredibly complex set of algorithms.

Please understand, this is not a question of “If.”

This is a question of “when.”

And when this accident takes place –– when an autonomous vehicle makes the logical decision to kill its own passengers, or a pedestrian, or someone in an oncoming car –– the repercussions of this accident (and the lawsuits that follow) will completely reshape the laws of this nation as they apply to personal injury, liability, and responsibility.

At the same time, the bedrock principals of cause and intent will soon share equal weight in the courtroom with examinations of process, purpose, and potential.  More specifically: what underlying decision making process is used by an autonomous vehicle when it purposely saves a life (or lives) while also extinguishing others, and what would be the future potential of the life (or lives) extinguished.

And all of this hinges on questions of liability and responsibility while, at the same time, fuzzying up the distinctions between legal liability and moral responsibility.

Who (or what) bears the blame for deaths that occur when a self-driving car makes the logic driven decision to kill someone?

Will car manufacturers shoulder most of the blame?

What about the software development teams who put together the autonomous car guidance coding?

What about the in-house ethics teams who –– as a committee –– create decision trees, outlining which deaths are morally acceptable, which deaths are totally unacceptable, and what sort of deaths occupy an amorphous gray area?

And what about hardware?  Will the original manufacturers of the various LIDAR and camera systems, providing self-driving cars with “vision,” also be brought in on the inevitable lawsuits

Lastly, what about the actual owner of the car?  What responsibility will he or she bear (provided they weren’t actually killed by their own car), when logic/software/algorithm driven deaths take place?

Given the pace of technological change –– and the acceptance of more and more semi-autonomous (and soon, fully autonomous) vehicles on our roadways –– these questions of liability and responsibility will soon be asked in a court of law.

Again, it’s just a matter of time before an autonomous vehicle makes the decision to kill someone.

Is our legal system ready to deal with the repercussions of the car’s decision?


Change is coming and –– brace yourself –– it’s coming fast

Though it might seem more-than-a-bit obvious to say this (and forgive me if this comes across sounding a bit cliched), but here in the 21st Century we are living through a period of profound social, financial, and technological transitions, and it’s all too easy to underestimate just how much change is taking place in the world –– not because the change is gradual, but because the change is fast, fluid, and next-to-impossible to keep up with.

One major technological shift that will bring about –– and IS bringing about –– foundation shaking changes to our world is the technology of self-driving cars (aka: “autonomous vehicles”).

Current estimates put the number of cars in the United States at roughly 250 million.

250 million cars, driven by humans.

250 million cars, that kill nearly 35,000 people per year.

250 million cars, that injure more than 5,000,000 people per year.

And all of this is about to change.

The first wave of this change is taking place in States like California, Nevada, Florida, and Michigan (along with the District of Columbia), where –– in February of 2017 –– self-driving cars have been granted the legal right of operation on roadways, streets, and highways.

And if trendlines are to be followed, we’ll see that legality extend to all 50 States within the next two to three years..

At the same time, as we see more and more autonomous vehicles make their way from the engineering labs to the dealerships and then out onto the highways, we’ll soon see a dramatic drop in the number of automobile related deaths and injuries taking place each year.

By their very design, utilizing full 360 degree views of their surroundings –– along with their networked and shared knowledge of all roadways and obstacles –– self-driving cars will bemarkedly safer than vehicles driven by humans.

But vehicle-related accidents and deaths will continue to take place.

Yes, there will be a markedly reduced number of accidents taking place, but accidents will still occur.

An in situations where an autonomous vehicle injures or kill its passengers –– or a pedestrian, or passengers in another car –– where does the blame lay?

Is the manufacturer of self-driving vehicle liable in any sort of way?

Or perhaps the engineers who developed the software allowing a car to be self-driving?

Or the owner of the autonomous vehicle –– do they shoulder any liability?

The simple fact is this: as self-driving vehicles become increasingly common on our streets, roads, and highways, the legal profession (along with federal and state legislatures) will itself go through a period of transition AND a period of confusion, where much of the confusion will be centered around the question of final liability.

Right now (in the early part of 2017), if you were to call up five different attorneys and ask for their opinions on liability issues related to autonomous vehicles, you will receive five completely different answers as to who/what/where the responsibility should lie in the case of an accident.

The point of all this –– and there really is a point to this whole post –– is that it’s important for lawyers, lawmakers, software developers, and car manufacturers to keep themselves up-to-date on all that is taking place in the world of autonomous technology and –– AND –– it’s even more important that lawyers, lawmakers, software developers, and car manufacturers are in open communication which each other. With all the changes taking place in our world, it’s far to easy to hunker down and confine oneself to a bubble of one’s own making. Just as self-driving cars will have open and shared knowledge of their surroundings (in other words, what one car knows all cars know), there needs to be an open and shared dialogue taking place amongst all the various parties involved with the development of self-driving vehicles AND the various parties who determine, legislate, and litigate self-driving vehicle.

The future will be here far sooner than expected.