What is Machine Learning and Why does Apple need it for their EV?

What is Machine Learning and Why does Apple need it for their EV? – Machine learning is a branch of artificial intelligence or artificial intelligence that allows systems to adapt to human abilities to learn. Without us realizing it, the use of machine learning is often present in everyday life. According to Forbes, machine learning is a current trend that will continue to grow in at least the next ten years.

What is Machine Learning and Why does Apple need it for their EV?

Before we know how important Machine Learning is to Apple in building their cars, we’ll first get to know what Machine Learning is.

What is Machine Learning?

Quoted from IBM, machine learning is a branch or application of artificial intelligence (artificial intelligence).

This science focuses on creating systems or algorithms that continuously learn from data and improve their accuracy over time without any particular programming.

In machine learning applications, algorithms or sequences of statistical processes are trained to find certain patterns and features in large amounts of data.

It aims to make a decision or prediction based on these data. The better the algorithm, the better the system’s decision and prediction accuracy will be.

Like humans who get smarter the more they learn, machines that process more and more data will produce more accurate output.

As previously mentioned, machine learning is now an important part of everyday activities. An example of the use of machine learning is a digital assistant that we can use on a smartphone to run a command.

In addition, machine learning applications can also be felt when advertisements on the internet recommend products that match our interests. The same applies to Netflix, which can find out preferences for movies or series according to what the user has watched so far.

Why is Machine Learning Important?

Machine Learning and Why does Apple need
© Pexels.com

According to Towards AI, machine learning is very important today.

Machine learning is useful for solving world problems in a scalable way.

The application of artificial intelligence can also be used in various industries and continues to be used by large industry owners and researchers so that they can continue to grow.

With machine learning, we can process and analyze larger and more complex data in less time.

In fact, according to The Wall Street Journal, machine learning and artificial intelligence have the potential to increase up to 16% or 13 trillion US dollars for the United States economy by 2030.

Of course, this will gradually affect the world economy as well.

Difference between Machine Learning and Artificial Intelligence

Machine Learning and Why does Apple need
© Freepik.com

You already know that machine learning is a branch of science from artificial intelligence or artificial intelligence.

Some of the key differences between machine learning and artificial intelligence are:

1. Success vs efficiency
The goal of artificial intelligence is to increase the chances of success, while machine learning aims to increase efficiency without being success-oriented.

2. Troubleshooting vs performance
Artificial intelligence aims to solve complex problems by simulating natural intelligence

Meanwhile, machine learning works by learning from data to improve the performance of a machine or system.

3. Decision making
Artificial intelligence simply works to make decisions. On the other hand, machine learning focuses on learning from input data.

4. Algorithm
Artificial intelligence mimics human capabilities in terms of response and behavior for systems. It is different with machine learning which is able to create its own algorithm for the learning process.

5. Optimization
Artificial intelligence is tasked with finding the optimal solution, while machine learning does not consider this.

Types of Machine Learning

Machine Learning and Why does Apple need
© Pexels.com

1. Supervised learning
Supervised machine learning is a machine learning algorithm using labeled data, i.e. input where the output is known. For example, a device may have data points labeled F (failed) or R (runs). Supervised learning algorithms accept a set of inputs with the right output. After that, this algorithm learns by comparing the actual output with the correct output to find errors or errors. In supervised learning, the algorithm can modify the model to fit the desired result. Typically, supervised learning is used in applications that predict future events based on historical data.

2. Semi-supervised learning
This machine learning method is not that different from supervised learning. However, semi-supervised learning uses labeled data and not to train algorithms. Usually, small amounts of labeled data are used and large amounts of unlabeled data are used. This machine learning method can be used with other methods such as classification, regression, and prediction. An example of using semi-supervised learning is to identify a person’s face on a webcam or smartphone camera.

3. Unsupervised learning
Unsupervised machine learning is the opposite of supervised learning. In this machine learning method, the processed data does not have a label and the system does not know the correct answer or output. The goal of machine learning with this method is to explore the data and find the structure in it. Usually, this method is used for transactional data.

For example, unsupervised learning can be used to identify consumer segments with similar attributes and group them so that they can be handled or treated the same in a digital marketing campaign. Not only that, supervised learning can also find the main attributes that distinguish between consumer segments.

4. Reinforcement Learning
Reinforcement learning is commonly used for robotics, game creation, and navigation. With this learning method, the algorithm will be able to find the action or treatment that produces the best output from the results of repeated trials (trial and error). There are three main components to reinforcement learning, namely the agent (decision maker), the environment (whatever the agent interacts with), and action (what the agent can do). The main purpose of machine learning reinforcement is for agents to determine what action maximizes payoff in a given time.

Why does ‘Apple Car’ need Machine Learning?

Apple is planning to use Machine Learning in their “Apple Car”, specifically because current processors are not fast enough to autonomously make key driving decisions.

A “machine learning system” for a vehicle is described in the patent, which uses ML to make predictions on how passengers will behave. So as not to rely too much on prediction algorithms and therefore be more dependent upon them, Apple instead focuses its efforts into the human-centric design of vehicles – combining aesthetics with practicality.

'Apple Car' need Machine Learning

The main goal of this project is to explore the design space for car-to-car communication systems. The team has created three different types of networks, each with varying degrees of complexity and performance that will be tested in a controlled environment. There are two scenarios outlined: “a fleet network,” which includes cars from all organizations within a company or industry; and “an individualized network.” In the latter scenario, one car receives data from its own local sensors but no other cars on their side roads.

This quote basically reinforces the point that you need to be quick and quick decisions can lead to a fatal outcome if not completed in time.

“Until relatively recently,” the patent says, “due to the limitations of hardware and software, it was impossible for a computer system to process certain data quickly enough in order for non-trivial navigation decisions to be made without human guidance.”

Recent progress in hardware and software development has been a big help for the average person, but we still have not achieved equality because of our unequal access to those technologies.

“Even with today’s fast processors, large memories, and advanced algorithms,” it continues, “however, the task of making timely and reasonable decisions… of the vehicle’s environment remains a significant challenge.”

The complexity of autonomous decision-making is that it “is based neither on excessively pessimistic assumptions nor on overly optimistic assumptions.” This means cars can drive themselves, but they won’t be able to do so without the help of other drivers. Other drivers in other cars also have unpredictable behaviors which might come into play.

So the real world is a lot messier than any test environment, meaning that autonomous driving decisions will have to be made even when there are “incomplete or noisy data.”

Over 17,000 words long, the patent describes situations to do with the car’s “action space.” It must make decisions within a certain distance and time.

'Apple Car' need Machine Learning

“In some states, such as when the vehicle is traveling on a largely-empty straight highway with no turns possible for several kilometers or miles,” continues the patent, “the number of actions to be evaluated may be relatively small; in other states, as when the vehicle approaches a crowded intersection, the number of actions may be much larger.”

In each case, the car’s systems have to determine “the current state of the environment” around the vehicle. Then it may need to identify “a corresponding set of feasible or proposed actions which can be undertaken.”

An action could be “turn left,” or “change lanes.” In at least some cases, ML can be used to help the car assign a number or value to each possible decision, and then determine the best course of action.

“[For example,] multiple instances or executions of a reinforcement learning model may be employed at the vehicle to obtain respective value metrics for the actions,” says the patent, “and the value metrics may be used to select the action to implement.”

This patent is credited to two inventors, Martin Levihn, and Pekka Tapani Raiko.

Apple Car Timeline until 2022

2008 — It would be the Apple Car’s CEO Steve Jobs who finally brings it to production. He is one of five people in the world with a special permit that allows him access to cars on private test tracks and other locations, as well as being able to put multiple prototypes through their paces at any given time. Some sources report that his first prototype was completed just days after he took charge back in 1997-1998, but others maintain this isn’t true because Bill Gates even had an idea for a similar product before then so if anything else happened it wasn’t much more than what we know today from Elon Musk. Either way, his name will forever go down in history alongside iconic carmakers like Henry Ford or anyone else you can think of!

2014 — This project is expected to be heavily backed by Apple and Tim Cook who, as of late, has been pro-environmental. The company will supposedly start production in 2020 with a goal of selling cars at $3500 USD each. This marks the first time that an electric car under its own brand line has come out without being distributed through Tesla Motors or another similar corporation. Read more: Codename ‘Titan’ for Apple’s Electric Car

2015 — Apple is said to have hired new employees for the project, alongside meeting with self-driving car experts and GoMentum Station, a California-based testing ground for autonomous vehicles. The firm hires Daimler Trucks subsidiary Torc to fit sensors to two Lexus SUVs in a project known internally as ‘Baja’. Read more: Apple’s self-driving cars head to reality since 2015.

2016 — Bloomberg reports that Apple has decided to switch its priority over to developing software to power self-driving solutions. The company is still looking towards creating a car of its own, but will not do it with the same resources and time as previously stated. Instead, Apple is now focusing on doing in house development so they can have more control over their product before releasing anything else into the wild.

2017 — Rumours have it that the technology will be ready for public use by 2020. The firm clearly has no plans to stop at just cars, as they are developing autonomous vans too. With this new development, Apple is moving towards taking a big step in car and transportation technologies. Read more: Apple’s program to develop autonomous vans. 

2018 — Apple registers 27 self-driving cars with California’s Department of Motor Vehicles. The firm enters a rumored partnership with Volkswagen to produce an autonomous electric shuttle bus. The FBI charged a former Apple employee for stealing trade secrets relating to project Titan. In August, an Apple self-driving car is rear-ended during road testing.

2019 — Drive.ai will continue to provide its self-driving technology for other automakers, while Apple’s acquisition of the company enables it to test and develop autonomous vehicle software on top of the hardware that they already produce. Read more: Apple Saves self-driving startup drive.ai

2020 — Apple has been working on the project for years and still hasn’t unveiled any of its plans to consumers. The fruit company is hoping their car will be one that’s not only practical in terms of fuel efficiency but also fun, offering a ride like no other automobile can provide. With rumors circling around Apple Car supposedly being built with sensors embedded into the bodywork, it seems as if this might actually become a reality and soon enough we’ll all be able to enjoy riding in an apple-shaped vehicle!

2021 — Industry scuttlebutt has it that Apple is in talks with Hyundai and Kia to jointly produce self-driving electric cars, but the vehicle manufacturers deny this. More rumours emerge of a 2024 launch goal — this time in collaboration with Toyota and its Korean partners.

2022 – For more information on the Apple car concept this year, click here!

Leave a Reply

Your email address will not be published.