Innovation That Matters

Top 7 AI Innovations From 2019

Innovation Snapshot

Without doubt, AI has been amongst the greatest sources of innovation in 2019, fueling transformation in most industries, from farming to filmmaking.

As artificial intelligence becomes more sophisticated, higher-skilled industries are also benefitting from its fruition. Robotic HR advisers, which use algorithms to manage portfolios, have proven extremely valuable during the hiring process. A Russian startup has created a chatbot named Vera to interview prospective employees. Likewise, US-based Harver is using machine learning to help reduce unconscious bias during the hiring process.  

Even medicine is not immune. Many sources have reported AI to be transforming the health industry. Recent advancements have demonstrated that AI systems are as good or better than radiologists at reading mammograms to identify cancer lesions.

With individuals from almost every industry being interested in the potential of AI and machine learning, here are seven of our favourite AI Innovations spotted during 2019. 


Photo source MIT Media Lab

Created by a 24-year-old PhD candidate from MIT, a device called AlterEgo can detect what you say, even when you are talking to yourself and not moving your mouth. It could be used to help those with memory problems or people who have lost the ability to speak.   

AlterEgo works by detecting the neuromuscular signals sent by the brain to the vocal cords. The device reads these signals and AI turns them into words. The speaker can hear the AI-generated responses through a tiny microphone that conducts sound through the bones of the skull and ear. Responses can also be read out loud through artificial voice technology.  

The device’s creator, Arnav Kapur, has been testing it on patients in hospitals and rehabilitation centres who have lost their ability to speak. Kapur has applied for a patent and has plans to develop AlterEgo into a commercial device.


Photo source Tim Gouw on Unsplash

Startup Harver created software that automates the hiring process, using data to sort, test and vet candidates. The program makes hiring easier, faster and decreases the chance of unconscious bias in the process, the company says. 

Harver is different from other AI-based hiring programs because it focuses on the pre-interview selection process. The software offers HR teams several adaptable assessment modules. The tests can examine everything from personality to typing skills. The program then assesses the results and determines which candidates are the best fit for an interview.

The software has already been used by several large, international companies like Netflix and Uber.


Photo source Arek Socha from Pixabay

Israel-based CyroReason created a machine-learning platform that can model a person’s immune system at a cellular level. The platform is the first of its kind and could lead to lower drug development costs and improve treatments’ effectiveness.

CyroReason’s platform identifies the “cause and effect” relationships on a cell level to determine how a person responds to medicine. That knowledge is crucial to learning how to treat that disease, according to the company.

The machine-learning model simulates the cells to “see” which ones can make a difference to treatment. The replicas are “cell-level blueprints of immune activity” within the context of disease and drug therapy, the company says.

“CytoReason is trying to build a computational model of human tissue,” co-founder David Harel said. The information learned from the model is used to support clinical trials of new drugs.


Photo source Anna Bernbaum

Vienna installed around 200 pedestrian crossing lights that can recognise when a person wants to cross the road. The system was commissioned by Municipal Department 33 of the City of Vienna and developed by a team at the Institute of Computer Graphics and Vision at TU Graz University. It is intended to replace the push-button system and can adapt to give large groups and people with disabilities more time to cross.

The system uses cameras mounted on the traffic light that has a large visual field. The research team used global movement models and recorded data to develop learning algorithms, which recognise when a pedestrian wants to cross the street. The system then triggers the light to change. Images are analysed locally by on-sight computers and are not saved.

The traffic lights are equipped with a monitoring system that can report faults immediately. They can also work in all types of light and weather conditions. The hope is that the system will not only make crossing safer and faster but will also allow smoother traffic flow.


Photo source Johny Goerend on Unsplash

Australia-based agtech startup FluroSat provides farmers with real-time information to assess plant health and detect crop stress. Its platform, FluroSense, combines satellite data, farming records and AI in an analytics engine to predict agricultural performance. Based on the data, it can provide recommendations on how to optimise crops.

It has been less than a year since FluroSense launched, yet it is already being used by more than 1000 agronomists across eight countries. In addition to their Sydney-base, the startup has also opened offices in Canberra, San Francisco and Kyiv, Ukraine.

FluroSense addresses one of the most basic issues with satellite data: how to use it efficiently and effectively. Unlimited imagery produces a lot of data, which can be hard to unpack. With the help of AI, FluroSense makes it easier for farmers to get the most out of the data and quickly make decisions.


Photo source Intelistyle

“Styled by AI” isn’t yet a commonly used term, although that may be about to change. London-based Intelistyle’s artificial intelligence (AI) chatbot stylist works with both retailers and customers. For retailers, the algorithm can “complete the look” by generating multiple outfits based around a single product and can recommend appropriate alternatives for out-of-stock items. With the app, the personal styling service can be accessed on any device, allowing customers a seamless move between online and offline shopping.

For shoppers, the chatbot recommends styles and outfits based on personal preference, body type and hair, eye colour and skin tone. Based on what is already in a shopper’s closet, it can recommend new buys as well as suggestions of combinations of items already owned. During 2019’s London Fashion Week, the outfit put together by Intelistyle’s algorithm was better received by a group of fashion experts than an outfit styled by a human.


Photo source Anna Bernbaum

Design engineering student Anna Bernbaum won a runner-up Dyson Award for her wearable device that can alert asthma sufferers to an upcoming attack. The device, dubbed Afflo, works by analysing the users’ respiration and the environment to determine possible triggers for an attack. The two streams of data are then analysed using a machine learning algorithm and the results are sent to the user via a mobile app. 

The goal is that once users know the type of situations and environments that tend to trigger asthma attacks or difficulty in breathing, they can take steps to limit their exposure. Data can also be reviewed remotely by medical professionals, allowing them to refine treatment plans more cost-effectively.

The prototype uses machine learning to differentiate between a cough and speech, correctly identifying a cough (a sign of asthma) with 82 per cent accuracy. Both the wearable and the sensor can pair with the patient’s phone via Bluetooth to transfer data.