Vertical Aerospace was established in 2016 by Stephen Fitzpatrick with a goal to create a decarbonising air travel using the best technology from the aviation, energy and automotive industries. Vertical Aerospace believes passionately in the power of electric flight to change the way the world travels because we shouldn’t sacrifice our planet to get from point A to point B. In 2017 a team of six engineers designed and build tha first demonstrator model, working closely with regulators and iterating the technology along the way. In 2018 their first eVTOL aircraft was granted flight permission by the Civil Aviation Authority.
One of Verticals flagship aircrafts, VA-1X, is a piloted all-electric vertical take-off and landing (eVTOL), which is capable of carrying four passengers for 100 miles at a cruise speed of 150 mph. Testing is scheduled for 2021 followed by certification in 2024. With the potential to be quite and economical these aircrafts have been touted as the next big thing in passenger aircrafts. Vertical Aerospace states that their aircraft will be heavily automated, but still both regulations and the public will require a pilot for years to come. VA-1X will automatically respond to an obsticte on the landing pad and will pull away from any collision. Using multiple propellers that point upwards VA-1X takes off and rotates those propellers horizontally in order to gain speed, it aims to carry four passengers and a pilot over short distances more cheaply than a helicopter.
Airlines operate within a framework of strict regulations, so it is yet unknown how a entirely new category will enter the market and the airspace. Vertical’s CEO says that they are already working pretty closely with Uk and European regulators.
In today’s world we are surrounded by a lot of devices which record and use our data. Usually when we think of data collection and recording we think of phones, home stations, tablets and others, but not cars. A large part of our personal data is being collected by our cars and can be used by police and criminals alike. Your car, depending on how new it is and the capabilities it has, could collect all sorts of data without your knowledge, including location data, at what point the doors were opened and your voice.
As an example I can bring up Joshua Wessel, who was charged with a murder based on the fact that the victims truck had a recording of Joshua’s voice at the time of the killing. A company called Berla Corp extracts data from vehicles on behalf of police. Berla’s software can identify unique ID’s of Bluetooth and Wi-Fi devices that were connected to the cars infotainment systems as well as text messages, call logs and contacts. And Berl’s software doesn’t stop there, it can look at the logs which were kept by the car’s internal computer, revealing location of some phone calls thanks to it’s built-in GPS or when specific doors were opened.
Car companies collect information on you whether it’s how far you drive, where you drive haw efficient your car is or what you listen to on the way to work. All of that data is valuable to understand who you are as a customer. Collection of data using your car started from a company called OnStar which was firstly introduced to make driving safer, you could not only call an ambulance in case of a crash, you could also use hands-free calling, turn-by-turn navigation and other emergency services. But later on this system evolved to track diagnostic, vehicle location and how often you use your vehicle and all this was collected in real time. On the website of OnStar they clearly mention that all the data collected could be used in marketing purposes, basically doing the same thing as Google. They mention that all of this information is presented to the driver, when buying an OnStar subscription, if the driver does not agree with the rules, his/her car will not be tracked, but most of the people don’t expect a car to collect their data so they don’t really read into the rules.
OnStar was one of the first to co;;etc personal data, but definitely not the last, nowadays all car companies have their own way of collecting personal data from drivers. “Connected vehicle” is a term describing a car which is connected to the internet, already 98% of new cars sold in the US and Europe are connected to the internet. All the collected data can be used to improve transportation, reduce emotions and traffic accidents and detecting crashes. One of the most valuable devices in a car is a E.D.R (event data recorder), it functions like a black box on a plane but for your car. E.D.R is triggered by accidents, sudden change in wheel speed or engine faults. Once it’s triggered, it records a wide range of elements that could be useful for crash investigators. Even though an E.D.R is located inside the car and store the data within the car, services like OnStar transmit data to off-site locations.
Most of the car companies don’t know how to consume the stream of data coming from their vehicles. A company called Otonomo takes all the raw data from your car, analyses it and sells it to third-party companies and splits profits with car manufacturers.
In my opinion it’s okay for navigation companies like Waze or insurance companies use that data to improve their services, but one of the biggest
See Tree is one of the leading companies in agricultural tech field. The number one priority of See Tree is to oversee their fruiting status of trees in order to increase their productivity. This calls for a device for automated detection of ripe fruits using machine learning techniques. To have better, faster results and to approach this task with all the tools that are needed, See Tree partnered with Ciklum, a company whose task is to help other companies in creating something new with their own specialists.
In this case Ciklum provided See Tree with a team of skilled R&D experts that had to find a solution to these challenges:
Small dataset (~500 annotated photos). It was impossible to collect fresh data because oranges were not in season.
High level of noise in dataset annotation. Annotations to some images didn’t contain required data. Some bounding boxes were not precise enough.
Dataset contained photo duplicates with different labelling.
High occlusion level in some pictures, made it difficult to separate instances.
To prepare data that is required for this task, Ciklum’s team duplicated photos and images of unlabled oranges or invalid polygons. Those images were filtered out of the dataset. The resulting dataset was divided into sets of train, validation and test, afterwards, stratified by the wide variety of oranges located on the picture.
The satisfactory end result became acquired with the Faster R-CNN architecture family.This pattern became a modern
method for resolving object detection task. ” Faster R-CNN is a two-stage object detection system in which the first
stage generates a sparse set of candidate object locations and the second stage classifies each candidate location as one of the foreground classes or as background using a convolutional neural network.”
Ciklum’s and See Tree’s prototype had these advantages:
able to detect ripe fruits with high accuracy, even if occluded;
although built to detect oranges, it can be easily re-trained to detect other kinds of fruits;