Companies across a wide range of industries and markets are assessing their ability to re-open safely in the coronavirus era. New rules and regulations, and new realities, apply. We have to think about how post-COVID will be different. Those on the vanguard of IT implementations are recognizing that artificial intelligence has vast potential.
Using the right AI can empower companies to be smarter about physical distancing, the preventative measure for COVID. It can also improve contact tracing that can provide key damage mitigation, decrease risk, and save lives. To the extent that companies can do these things effectively, they are making everyone, the employees, the customers, and the suppliers, safer.
It’s almost as if artificial intelligence was made for this kind of scenario.
The automated tracking and reporting and monitoring that AI programs do is a vital resource for today’s coronavirus age. Just as the Internet offers a key way to sell goods and services and complete other transactions while social distancing, AI provides a real way to leverage smart tech to make us even safer.
Benefits of AI implementation for COVID-19
Companies that can successfully work with vendors and consultants to implement AI/ML solutions like these will benefit in some key ways.
First, they will provide demonstrated compliance with a range of standards, which can open up many doors for a business.
State and federal leaders may be more likely to allow a particular business to expand its model if it can prove that it has these systems in place to effectively contain or track COVID. Proving enforcement is demonstrated compliance, which means a great deal in today’s world. (This resource from Deloitte covers a broad waterfront. Didn’t know there were so many aspects of COVID compliance? Take a closer look.)
There’s also the labor cost of handling things manually. Obvious limitations apply to manual counts in stores and other facilities. This requires staffers to stand at a door or other key assessment point continually, and tally up numbers in their head! This is a process that computers are much better at, inherently, and a process that computers can do much more cheaply and efficiently.
A corollary benefit is that these automated systems take the human error out of the equation. It’s easy to imagine one of these hard-working counters simply making a counting mistake, or failing to register a particular individual’s motions. An AI program, on the other hand, is a never-flagging sentinel, one that never needs a break or makes a counting mistake, or takes its electronic eye off of the ball.
Specialized Solutions for Physical Distancing
Artificial intelligence systems for physical distancing will often use one of two modalities – either a haptics approach, which leverages a tactile user sensation, or a set of cameras and sensors.
Some companies favor the haptics systems because the buzzing or touch implementation of devices shows the end-user what’s happening around them. However, camera count systems can be entirely effective in controlling traffic inside of a facility (even last year, Tech Times was surveying the best people counting tools of this kind.)
It’s important to note that many of these solutions existed even before COVID struck. Businesses have been using them for sophisticated point of sale monitoring, for store movements, or for the drive-through patterns, to see how customers move, in order to glean insights. Now, the same kinds of data are suddenly valuable to public safety efforts.
Specialized Solutions for Contact Tracing
The AI solutions for contact tracing have much more to do with data implementations. Here’s MIT Technology Review’s coverage of Google systems being rolled out everywhere.
Essentially, these systems will take the automated data around an individual’s location history, and apply it to a theoretical infection. When someone reports (or is reported) infected, those location data will be used to evaluate and backtrack to determine the contact tracing results.
With these implementations, it becomes critically important to talk about privacy issues. These systems will require that user data is kept secure, and will also require buy-in from end-users themselves, which simply can’t be overlooked.
Here are two key methods that are very instructive in addressing both buy-in and privacy. The first is an opt-in and opt-out system. Opt-in, opt-out ensures that the company has recorded that an end-user has assented to use of the technology. Some of the concerns around feasibility studies may have to do with what percentage of users will or won’t opt-in.
The other key pillar here is to anonymize data. If instead of identifying an individual, the contact tracing simply shows that a given anonymous point 1 was contacted with anonymous points 2 and 3, that does not require the same data protections that identified contacts do.
Key Functionality with AI Systems
One reason that businesses (as well as schools and universities) are building their own proprietary AI systems for physical distancing is that these systems can be used to isolate COVID infections.
Another key piece of functionality here, as addressed above, is keeping crowd sizes low. This takes the guesswork out of complying with state directives to this effect. Suppose a college wants to ascertain how many students are grouping on the quad, or inside a particular lecture hall. The tools make this openly transparent to support proactive administrative goals.
All of this information can also be reported back to a command center or an IT department for further use and longer-term planning. Again, the goal is to keep people safe from the virus. Schools or businesses may very well use the longer-term insights to craft articles in a newsletter for public awareness, to set policy, or to call the shots on which planned events will meet safety criteria.
The above shows how AI can work to a firm’s post-COVID advantage. By thinking about these capabilities and applications, we can envision what safety and compliance will look like in the months and years to come.