Using AI to Support Experts

“Artificial intelligence can boost our analytical and decision-making abilities by providing the right information at the right time.”

Collaborative Intelligence: Humans and AI Are Joining Forces, Harvard Business Review, July 2018.

Whether applied to agriculture, healthcare, or communications, technology supports and enhances a human’s task or experience. Utilizing computing power to process vast sums of data, technology can contribute huge analysis and insights in a timeframe virtually impossible to humans, allowing experts to identify and focus on priority cases. Here we look at cases and how this machine learning approach is in use today.

AI cases in review

There are exciting examples of AI and advanced analytical tools being used to enhance productivity and efficiency across a wide cross-section of industries. Here let’s look at two examples:

1. Healthcare:

As part of a five-year partnership between Moorfields Eye Hospital NHS Foundation Trust and DeepMind Health, researchers from Moorfields and the UCL Institute of Ophthalmology successfully used machine learning to identify signs of eye disease and make an appropriate referral. Published on Nature Medicine's website, this was achieved through using technology to analyze thousands of historic eye scans to identify and learn from patterns and algorithms. With a 94% accuracy in referral decisions (matching world-leading eye experts), the technology is deemed to be the future of eye testing, enabling earlier diagnosis and more accurate prioritizing of patients.

2. Finance

Some of the largest fraud losses are suffered in the telecoms industry, with an estimated annual loss of almost $30 billion, according to the most recent estimations by the Communications Fraud Control Association (CFCA). The faster fraud can be detected, the quicker it can be shut down, minimizing financial and reputational losses. Fraud detection software can process and detect complex patterns in vast amounts of data to identify abnormalities or suspicious data that require additional investigation. This releases investigators from data trawling to allow them to focus on qualified cases, providing the opportunity to detect fraud early and minimize damage, financial or otherwise.

AI cases in Senseye PdM

Cases created automatically in Senseye PdM, the leading product from Senseye for Predictive Maintenance (PdM) provides all the information to carry out an investigation into a possible future asset failure, allowing it to be fixed before it becomes an issue and disturbs the production line. Senseye PdM’s technology capabilities are used to monitor massive volumes of assets, perform calculations relating to an issue and present users with a case for investigation. The case is made up of details of the potential problem, including what led to a potential issue being picked up, what the concern is and why it should be investigated. Scalability: the silver lining.

AI has the ability to process data and perform calculations faster and more accurately than humans, but it is the availability of hosting this data in the cloud that is enabling the other real big opportunity: scalability. For humans, increased data would mean a recruitment exercise, but with cloud computing you can just scale up almost instantly when required. Also, just as humans learn from experience, so does AI.  When the scope of an AI program’s remit is expanded, so is the ability to learn, refine and improve outcomes.

Case closed

Whether it’s Alexa or Siri giving you the current train times and weather forecast, your bank freezing your account for  unusual payment behavior to protect your assets,  your sight being assessed by a machine, or Senseye PdM creating a case to alert of unusual machine behavior , we are surrounded by technology aiming to better protect and serve us.


Download the white paper