How AI or machine learning can improve quality assurance: six tips

Moreover, poor quality can become dangerous if faults cause injuries or other complications. Fortunately, no matter what kind of company you have, artificial intelligence or AI may assist with your quality assurance needs.

Here are seven examples, along with AI and QA tips you could adopt for your organisation.

1. Prioritising test cases

Quality assurance involves a number of different phases and tasks, which all link together to ensure that errors and defects are eradicated before they get to the customer. For this to be properly executed however, staff need to prioritise which duties need to be done immediately, and which can wait. This ensures that vigilance is well-balanced with timeliness across the QA cycle.

Factors that typically need to be taken into account include code coverage, risk analysis, and business criticality. AI-based selection tools are available that can be utilised to help staff determine where to start. To achieve this, such tools can analyse code and documents and behaviours.

By using AI to prioritise test cases in quality assurance, users can make sure that no gaps in the process are present, and that all members of staff are on the same page regarding progress and what needs to be done.

2. Helping developers release error-free software

One of the areas where AI is proving its worth for quality assurance is in the software development sector. AI seems particularly well-suited to regression testing.

That approach requires checking to ensure previously tested versions of software keep working as expected following code modifications. Or, AI could help create new test cases. Some AI models can recognise or come up with scenarios without prior exposure to them.

If you’re thinking about using AI for testing help, confirm which processes that typically take humans the longest to do or where the errors happen most often. Then, assess whether AI might avoid some of those issues and speed up the steps testers typically go through when verifying all is well with new software.

Also, keep in mind that using AI for software testing works best when you have a large data set. That’s why training your AI models thoroughly is so necessary, and not a step to take hastily.

3. Reducing instances of foodborne illness

According to data from the U.S. Centers for Disease Control and Prevention, foodborne illnesses make 48 million people in the United States sick each year and send 128,000 to the hospital. Companies that manufacture any consumable product must uphold strict practices in their facilities to prevent oversights that could lead to food poisoning and recalls.

A team at Google worked with Harvard researchers to build an AI model that can reportedly identify possible food safety issues in near real-time. It looks at Google search queries from users that have their location data-enabled.

For example, if a person searched for things consistent with food poisoning, like “causes for diarrhea,” the AI pulls the location data to see which restaurants the person visited recently. Then, health departments received lists of which establishments might have issues, and inspectors went to investigate.

Although Google only used it in a couple of cities, the AI was better at finding problems than a previous system where people complained about restaurants’ food safety shortcomings. More specifically, the AI model found unsafe restaurants 52.1 per cent of the time, compared to 39.4 per cent with the complaint-based method.

Your company may not have the resources Harvard and Google do. However, there is a good chance that you still may find AI helps. Some tools track sentiment analysis on social media. If you realise bad feedback suddenly pours in overnight or within a few hours, you need to look into the matter — regardless of what you sell.

4. Letting humans focus on other tasks

At many companies, quality assurance is only one duty a person assumes during a workday. For example, journalists have many tasks that come together to support quality assurance. They have to check their sources, follow their organisation’s style guide and screen for typos and grammatical errors when producing their content.

Some people believe AI will become instrumental in allowing journalists and their editors to devote more time to other parts of their workflow, such as digging into a story and looking for more people willing to give interviews. They assert AI is not taking jobs away, but changing how people work and allowing them to concentrate more on the responsibilities technology can’t do as well.

If you think AI could support your QA goals by taking care of some of the time-intensive tasks humans do at your organisation, look for purposeful and easy-to-use solutions. Also, it’s helpful to use metrics that confirm the AI genuinely does save people time without sacrificing quality.

5. Detecting defects before products reach the market

Tighter markets and higher global competition are some of the things that caused manufacturers to emphasise quality assurance more than ever. A company that excels in QA delivers on their promises to keep customers satisfied. They typically build QA into all aspects of their business. In the manufacturing sector, it’s easy to see how expensive QA problems could become if left unaddressed.

Many automotive companies use the Industrial Internet of Things sensors regarding AI and quality assurance. The sensors gather data, and AI algorithms analyse it to pinpoint possible issues.

Perhaps the most beneficial aspect of this method is that manufacturers can become aware of problems earlier. Then, they can significantly decrease the likelihood of cars coming off the assembly line and getting sold to customers with defective parts.

When your company wants to reduce the number of product defects, consider whether other technologies, such as sensors or big data interfaces, could equip you to learn where the problems start and why they happen. That way, it should be easier to discover where the gaps in your QA process exist, and how you should deal with them.

6. Enhancing and personalising medical care

Medical facility executives gauge a variety of statistics to track patient outcomes and find the problems that may lead to substandard medical treatment. For example, they may look at hospital readmission rates, the average length of a hospital stay, how long people wait before getting seen in emergency rooms and so on.

Those measurements are all part of seeing how a facility’s level of quality stacks up to competitors. Florida’s Flagler Hospital uses AI tools to aid in care for patients with high-mortality conditions. The technology cuts costs and hospital stay lengths. For example, the tool can save over $1,300 in direct variable costs for pneumonia patients and cut their stays by two days.

You can use a similar approach by figuring out whether AI could improve any time or expense-based statistics. As always, understand quality levels should stay high, even when other measurements decline.

7. Providing better educational outcomes

The training employees receive also connects to QA. Researchers at the National University of Singapore developed a tool that uses AI to customise training content to individual users.

They say this method improves cognition more than other training techniques. One of the things the AI tech showed is that people respond to training intensity differently, and a customised approach may promote comprehension.

One way you might use AI and quality assurance for training is by keeping track of what modules people covered before and how they responded to different content delivery styles, such as video or audio.