Automotive Manufacturing

Challenge

Our client, nervous about letting faulty cars through their production line, was detecting production faults with a very high false positive rate. It takes a long time and a lot of resources to check all vehicles; therefore, if you can be more discerning about which ones are not actually defective, that would signify a significant OPEX reduction.

Data

250-300k rows of data from multiple sensor readings.

Process

Logical Glue built Neural Network models to recognise a number of different specific faults using interactions within the sensor data, rather than a catch-all generic fault.

Outcome

By bucketing the results of these models according to the level of fault risk, it was possible to greatly increase the number of cars that cleared production testing without letting any more faulty cars through. This allowed the manufacturer to free up significant resource from the checking process.

Healthcare / Pharmaceutical

Challenge

Our client needed a means of predicting drug adherence, so they could direct resources to patients at risk of stopping their treatment early.

Data

Wide sources of data around characteristics of the person, such as type of drug, method of administration, etc., with a sample size of ~100k.

Process

Logical Glue developed an XAI model that created rules spanning the different data sources, meaning that they would have been unlikely to be discovered by a human expert. Running these rules on an individual person allowed them to be assessed for adherence risk.

Outcome

Applying the rules generated by the model established a group of people least likely to adhere, allowing targeted early intervention with a much better success rate than the existing untargeted process used by pharmaceutical companies.

E-commerce and Marketing

Challenge

An online retailer wished to optimise their Google keywords coverage in order to maximise their click through and conversion rates.

Data

A set of 1m Google searches by users, including key words, and click-through and conversion rate data.

Process

Logical Glue built an XAI model that was able to determine high-level combinations of keyword types that led to strong conversion rates. This allowed the creation of type classifications for each keyword in a customer search, meaning a “conversion” score for different keyword combinations could be determined.

Outcome

The model allowed the client to realise price savings by specifying previously unused keyword combinations that gave a high probability of customer conversion.

Get in touch

Let Logical Glue build you a brighter business future.

Logical Glue’s team of highly experienced machine learning experts can help you to harness the explosive power behind XAI and Data Science: our cost-effective solution provides business-critical assurance through our expert marrying of cloud-based machine learning, data science and lending acumen as an intuitive learning experience.

Let's Talk