Tag: autonomous vehicles

Progress of Autonomous Vehicles Over Time

Driving a car is the MOST dangerous thing we do every day – 40,000 people die in cars each year.

Humans are really bad drivers.

To get a driver’s license, you’re given a 25 question multiple choice test at the DMV and then get behind the wheel.

Humans don’t work towards being excellent drivers the way they train for a marathon, study for medical school, or practice an instrument.

Poor driving is amplified by distraction – checking phone notifications, texts, social feed, etc. How much can the average person be expected to maintain focus with their eyes off the road?

Autonomous vehicles could save tens of thousands of human lives per year.

Get the Future of Technology letter each month. Sign up below.

Processing…
Success! You're on the list.

The image below shows a graph of the advancement and sophistication of autonomous vehicles as time and technology moves forward.

Progress of AV’s hits an inflection point where quick progress displays itself as a steep learning curve, slowly approaching an asymptotic limit of perfectly flawless autonomous driving.

As advancement of AV’s approaches this limit, theoretically, system will never be perfect, but it will surely become good enough that the chances of a collision by an autonomous vehicle with another object is practically zero.

This is the point at which vehicles are 99.9999…..% safe, highlighted as the green line. As systems continue to be developed, we expect the “march of nines” above to approach closer and closer to 100%.

There is a big difference between a small fender bender and a fatal collision.

Before self driving technology progress reaches a point where it is statistically unlikely enough that a collision will ever occur on a road, an autonomous system will need to reach a point where the risk of fatal collision is practically zero.

This may be mitigated by incorporating risk avoidance technologies such as slowing down in high-traffic areas, or even designing fleets of cars that are able to communicate from one to the other.

Although autonomy progress has been drawn as a sigmoidal curve above, there may be an argument that the actual progress would look more like a logarithmic curve, if there were no time of slow progress before the inflection point.

In either case, self driving cars continue get better. Some companies have already built autonomous vehicles that feel safer than human driven cars. But these systems are still not entirely ready for the roads.

Humans do not accept AVs unless they are 100% safe, even if they are safer than human drivers. It’s not that humans feel nonchalantly towards the 40,000 people that dies in car crashes each year, it more that humans have extremely high expectations of technology.

Despite the fact that our cell phones send information through the air, we get frustrated when internet speeds are slow and it takes us a few seconds longer to get an answer from Google. A lack of complacency isn’t exactly a negative thing, it promotes technological advancement.

Any non-zero number of self driving car crash fatalities is absolutely unacceptable. The infamous av-Uber crash in Phoenix, Arizona was a tragic nightmare. Ultimately, autonomous vehicle technology must be perfect before humans will accept it.

Does this highlight some principle that is distinct to human mentality? Here are two examples:

Example 1: Humans have irrational fears

I have a few close friends who choose not to surf or go in the ocean because they are afraid of sharks. This is socially acceptable. But I have never met a single person that avoids riding in an automobile out of fear.

To be fair, transportation is pretty much mandatory for a lot of things in life, whereas going swimming in the ocean is trivial and not a requirement.

Why does a fear of sharks continue to be so disproportionally high among humans, compared to driving, which is orders of magnitude more dangerous?

Example #2: Humans are borderline incompetent at most things…

And our only hope is to create tools to help us accomplish the things we need to do. Expecting a human to drive a car is like expecting someone to prepare and serve a full course dinner without any of the tools that exist in a kitchen. While it is surely possible that someone might be able to build a fire without matches and maintain a consistent temperature with which to cook their food, it is extremely likely that they burn the food and making an awful tasting meal. Without tools that help us cook, we’re incompetent. With tools like utensils and appliances, most people still have a hard time successfully preparing a meal. Even with the most advanced stovetop and cookware, cooking is difficult and takes just the right amount of time and patience to get right.

Transportation is no different. Humans were incompetent at all forms of transportation before railroads and the combustion engine. With engines and automobiles, we’re still awful drivers.

Lewis Hamilton (Mercedes) - GP of Spain 2019
source: Eurosport

A car is just a tool. It is a solution to the slow transportation problem.

Some people drive cars for fun. Most people drive cars because they have the human centric need to move around from one place to another at their free will. Autonomous vehicles will make transportation more safe and effective.

Get the Future of Technology letter each month. Sign up below.

Processing…
Success! You're on the list.

Source:

  1. According to the National Safety Council, over 40,000 people were killed in vehicle-related incidents in 2018. During the previous 3 years, there were more than 120,000 total fatalities.

Tesla Accelerates Sustainable Energy

How will Tesla enable sustainable transportation?

disclaimer: written by a TSLA shareholder. Opinion. Not investment advice. Do your own research. All information here comes from publicly available sources or is speculation / guessing. Please fact check this blog post. (going overboard on the disclaimer after a particularly funny reddit comment).

Tesla has distinguished themselves as a company that builds both software as well as physical products and hardware.

One of the few companies whose mission appears to be sustainability over profit, they continue to innovate and create the best technology, forcing other players in the market to try to keep up.

Tesla’s Big Goals:

Tesla’s number one mission is to accelerate the transition to sustainable energy. Tesla is progressing in a few main areas to achieve this goal:

  1. produce more affordable electric vehicles
  2. build systems for energy storage
  3. be the best at manufacturing

The CEO of Tesla, Elon Musk, has stated how he believes “you have to have a goal”. Following his earlier statement, Elon kept his word – the company’s goals were concisely outlined during the Tesla battery day event this past September.

Get the Future of Technology letter each month. Sign up below.

Processing…
Success! You're on the list.

Goal 1: Tesla building Affordable Electric Vehicles

Today, less than 1% of the cars on earth (the “global fleet”) are electric. To increase this, Tesla will build a car that anyone can afford. Tesla has announced plans to build a $25,000 electric car.

With what started as a luxury, high-end car, Tesla is working towards reaching economies of scale to move towards high volume production of a car for the mass market.

Once Tesla reaches full-scale manufacturing and production, Tesla wants to produce and sell 20 million cars per year, enough to replace 1% of the gasoline cars with electric vehicles.

To achieve these numbers, Tesla must increase production of their cars by 40X compared to their 2020 manufacturing numbers.

The center controls are completely touch-screen. source: Tesla

Elon has hypothesized that internal combustion engine industry WILL NOT EXIST in the future, aside from perhaps in museums and hobbyists.

Autonomy & self-driving cars:

Creating self-driving cars is not directly related to reducing carbon emissions, yet it does provide a few solutions that will make buying a Tesla an extremely attractive purchase:

  1. Safety. Autopilot, designed to avoid collisions, has the potential to save the lives of at least 40,000 people per year that die in automobile fatalities. With traditional automobiles, accident probability is 2.1 collisions per million miles. With Tesla autopilot, the probability is 0.3 collisions per million miles.
  2. Entertainment will be important in the car once human attention is no longer just being used to drive. This could include video games (the vehicles already have a number of games available), socialization, reading, working, etc.

It is hard to say just how valuable autonomy is to each customer, but the parameters are as follows: Since autonomy is valuable to cars at an individual level, the value of autonomy for Tesla = value of each car * value of autonomy per vehicle.

Full-self-driving technology BETA version feels like it is close to being released to the market after seeing a few of the CEO’s recent tweets on the subject.

source: Twitter

If the FSV features help the company sell more cars, then the company is that much closer to achieving its mission of an automobile economy based on sustainable energy.

Goal 2: Energy Storage – Tesla Batteries, etc.

To be truly sustainable, energy must be accessible and affordable to everyone. Tesla plans to make vehicles and grid batteries that cost less. This includes reducing the cost of energy per kilowatt hour by one half.

Tesla is working to change the trajectory of the curve of cost per kilowatt hour of energy, pushing the cost lower, shown as the red line. source: Tesla Battery day.

At Tesla battery day, the heads of the company mentioned that reducing the cost per kilowatt hour of batteries is not happening fast enough. This was demonstrated by showing the curve of cost per kilowatt hour of batteries and the slow rate of improvement. Its flattening out, as shown in the photo above.

Battery design:

Tesla is al re-engineering the battery cell design, manufacturing, and production processes to create more affordable cells.

Tab-less batteries. source: Tesla Battery Day
  • Tesla batteries are cylindrical, and newer versions are larger (bigger cylinder cells cost less)
  • Tab length in batteries: older batteries had tabs located at anode and cathode ends, which added to the distance an electron has to travel through a battery. Tesla got rid of tabs, so the electron only has to travel a shorter distance, making them more efficient.
  • Battery filler is not only flame retardant, and it is a structural adhesive. Glues cells to the top and bottom of the sheet.
  • Battery cells are load-bearing, made of steel, dual-purpose as the structure of the car itself. (see below)
The image shows how larger battery cells (blue cylinders) in the bottom image are more efficiently packed into the car. Older design at the top has a large amount of wasted space, shown in red. source: Tesla Battery Day
  • Anode: Tesla uses silicon instead of graphite (graphite is carbon based) for the anode. Silicon is the 2nd most abundant element on Earth, present in Earth’s crust as silicon dioxide, commonly known as sand). Stores 9x more lithium than graphite. Problem with silicon is that it expands in the cells. They use raw metallurgical silicon and design batteries to be able account for expansion.
  • Cathode: the cathode holds the lithium and retains its structure. Nickel is the cheapest and has the highest energy density, but Nickel presents challenges with chemical stability. Cobalt is more expensive, yet more stable than Nickel. For the most energy intensive batteries (like the semi-truck or the cyber-truck) they will use full nickel. The goal is maximizing nickel and gradually removing cobalt from battery manufacturing. The company has added coatings and dopants to stabilize nickel in the batteries. Cathode materials are purchased and priced based on the London metal exchange (LME).
  • Lithium: lithium is plentiful in the US, Tesla already has access to enough for every car (once they are building 20M+ per year). The company mines clay containing lithium in areas of the US where the ground has high concentrations. They extract the lithium via an environmentally friendly process involving table salt NaCl. After mixing it with salt and water, the lithium is extracted because lithium bonds extremely strongly to Cl-. The lithium effectively knocks off the sodium atoms, and we are left with LiCl salt, which the company can use for their battery manufacturing.
  • The company will eventually recycle materials in the used batteries to make new batteries.

Goal 3: Manufacturing

In addition to the number one goal of accelerating the adoption of sustainable energy, Tesla wants to be the best at manufacturing. Elon stated Tesla needs to be “better than anyone at manufacturing”. The company has created a vertically-integrated car from the ground up. They build everything in house, outsourcing little of the process.

The remarkable thing being built by Tesla is actually not the car, but how. The way the company built the car, with heavily automated robotic factories is impressive.

Tesla is building 4 types of products for consumers:

  • Energy generation (solar panels / solar roof)
  • Energy storage (Tesla Powercell) – customers want the freedom to charge at home. The Tesla Powercell product allows people to do so.
  • Electric vehicles (cars and trucks)
  • Automated factories. The company has engineered machines to build the car, supporting the creation of each product. While these three products are very much in the foreground, the importance of the robotic factory in the background has given Tesla a wide competitive advantage that will be extremely difficult to copy.

Factories

  • Sustainable factories include car factories built with solar.
  • Factory close to consumers (on each continent) shortens the supply chain, quicker delivery to customers. Factory in Fremont California, Nevada, Austin TX, Berlin, Shanghai China.
    • Tesla is the only American car company with manufacturing facilities in China.
  • Largest casting machine ever to make the front and rear casting in one piece.
source: Tesla Battery Day

Engineering

The ability to do more with less is an important strategy in engineering for elegance.

The company focuses more on metrics within the context of product improvements and manufacturing than purely financial areas. As of Battery day, the company reported:

  • Reducing number of parts in the car: now 370 fewer parts.
  • Reducing floorspace required in the factory by 35%.
    • Ensuring each cubic meter of the factory floor does useful work.
  • Stop using cobalt in batteries, mainly use nickel now.
  • Materials engineering the frame of the car: Developed their own high-strength casting alloy of aluminum that does not require coating or heat treatment. (Heat treatment historically causes alloys to lose shape)
  • Shortening the supply chain for resources: Reducing miles traveled by materials that end up in cathode by 80%.
  • 10% mass reduction in the car.
  • 14% range increase
  • “Electric energy costs are half those of diesel. With fewer systems to maintain, the Tesla Semi provides $200,000+ in fuel savings and a two-year payback period.” – Tesla.com
Tesla Semi-truck rendering source: Tesla

Get the Future of Technology letter each month. Sign up below.

Processing…
Success! You're on the list.

Sources:

  1. Tesla Battery Day presentation Deck: https://tesla-share.thron.com/content/?id=96ea71cf-8fda-4648-a62c-753af436c3b6&pkey=S1dbei4
  2. http://www.ev-volumes.com/
  3. IHS
  4. OICA
  5. https://www.tesla.com/blog/secret-tesla-motors-master-plan-just-between-you-and-me
  6. https://www.tesla.com/blog/master-plan-part-deux
  7. Tesla website
  8. TSLA 10k / annual report
  9. twitter