our solution
about us


Tel: +49 (0)30 40363399

Feedback first: developing the Machine Insight Center with users in mind and the minds of our users

In developing the new Machine Insight Center software for the AiSight predictive maintenance solution, we had the advantage of a great resource: our dashboard users. Read more to learn how having access to feedback made a difference in our development process.

AiSight's Machine Insight Center is the all-new software for our predictive maintenance solution. We want this to be the best user interface possible. That’s why we’re collecting feedback from maintenance and production professionals who use our solution—this feedback is at the heart of our development process.

This goes back to the earliest stages of the Machine Insight Center’s development. Thanks to user feedback, we knew that we’d need to build new software from the ground up.

If you missed the introduction to this series, you can read more about that decision here.

It's also thanks to user feedback that we knew what we’d need to implement in the Machine Insight Center: features like machine histories, visualized alerts, more flexible organization options, and user permissions. And because we did so a complete reconstruction, we were able to incorporate those features into a faster, more intuitive, and more scalable piece of software.

We were lucky to be in this position! We can’t overrate communication with our customers—learning what really works for our users is invaluable. We know this thanks to our experience developing both the Machine Insight Center and the old AiSight dashboard. These were two different development journeys.

Back when we developed the original AiSight dashboard, we did so without the benefit of direct customer feedback. We just didn’t have the ability to communicate that we have now. That shortcoming led us to develop the Machine Insight Center with feedback in mind—a change in both our development approach and outcome.

Developing the original AiSight predictive maintenance dashboard

We developed the original AiSight dashboard because our solution needed it—a complete predictive maintenance solution has to include a software component that delivers insights to users.

And our team had the expertise to pull it off. Our founder Matthias Auf der Mauer has a storied history with sensors, including time working as head of IoT at a major automaker. Since founding AiSight, he's stacked the team with vibration experts. Their combined knowledge and competencies provided bedrock information upon which to build the dashboard.

But until we launched, their expertise was the only input we had to design the dashboard. Out of pure necessity, we'd prioritized delivering a product over discovering what the product could be. It turned out, fulfilling that necessity was a critical step to furthering discovery—and development—of our software.

Since launching the dashboard, we've had an invaluable new resource: feedback from users. With this information in hand, and our team committed to developing a new app from scratch, we could change our development approach from one prioritizing delivery to one of continuous discovery.

Continuous discovery for continuous improvement

We keep AiSight connected to the factory floor—not just through our sensors, but through training and company culture. No matter how removed the daily tasks of any member of our team may be from maintenance and production, we want them to have a hands-on understanding of what we do. That's why we teach every new joiner how to install sensors, and always send them on at least one installation at a customers’ site.

The upshot is not only connecting team members to what we do, but providing information on how we can make installing our solution easier. Thus, we gain steady feedback on how to improve one facet of our hardware. To get the same kind of feedback on our software, we'd need to tap a stream of steady feedback from our software users.

With the dashboard in service, that stream existed. The early feedback we got from dashboard users was invaluable, but we had no systematic process for collecting it. We relied instead on the traditional conduits of information between customers and our software team—our teammates in sales and customer success. And we needed more. Furthermore, we needed to connect with the busy maintenance and production professionals who actually worked with the dashboard. It wasn't a given that they would have direct contact with us, or time to spare.

Going into development of the Machine Insight Center, we knew that we’d need to open channels for constant communication with users. This would inform us not only of which features they'd like, but why they wanted them and what problems they wanted to solve. This would, in turn, create a platform for delivering solutions to real problems, not just features on an abstract wish list.

With a steady flow of information, we could develop the Machine Insight Center with continuous discovery in mind, and with the outcome of our development in view—wherever we were in the development process.

How to secure testers for continuous discovery

To design from a user-centric and data-informed perspective, we'd need both users and data. That meant recruiting testers.

Our first attempt at recruiting was less than successful. We went through an agency, who found us testers from the maintenance industry. But their feedback was not as precise and actionable as we'd hoped.

Nevertheless, we were able to develop a minimum viable product. To test it, we assembled an advisory board of existing clients, and launched a private beta. Then we waited.

And waited.

We did get quality feedback from the private beta—it just wasn’t enough. We wanted weekly discovery, and our elite team of testers was too busy producing and maintaining to check in that often.

After delivering on the feedback we did receive, we decided to open a public beta. This opened the Machine Insight Center up for all of our users to test and provide feedback. As more users switched to the Machine Insight Center, we gained a trickle of new feedback, which soon became a stream.

We now have the quantity and quality of feedback that we need for continuous discovery—and we always want more. That's why you too can contribute to adapting the Machine Insight Center to your workflow and specific needs.

See how the AiSight Machine Insight Center can deliver intuitive predictive maintenance insights to your plant.

In the next issue of the Machine Insight Center development story, we'll look at how we manage the feedback we collect.