The increasing availability of powerful Artificial Intelligence (AI) models is heralding a full-fledged fourth industrial revolution. Low threshold yet highly flexible solutions such as ChatGPT are pushing all industries to make their move.

And move they will. According to Accenture, 84% of the executives claim that in their organization growth objectives can’t be reached without scaling AI, and 76% of them struggle to scale the AI across their organization. In addition, over 75% of executives believe it’s a “do or die” situation, and they have no more than five years to act.

It is hard to disagree. With an estimated 300% return on AI investments and significantly increased operational efficiency, organizations betting on AI are bound to record tangible benefits. Unfortunately, only 20% of all companies have leaped so far, while the vast majority still need to get started.

After all, transforming into an AI-powered organization is more complex than replacing a few components here and there with new tech. A successful transformation will require alignment with business strategy, dedicated multidisciplinary teams, and advanced data management capabilities.

According to Accenture, only 5% of organizations have industrialized AI to support their growth. 20% have scaled AI strategically, and 80% still struggle to utilize AI.

Improving efficiency with smart automation

There’s no path that fits all, either. Every organization will have to define its distinct AI journey to be successful.

Recently we helped one of our long-term customers on their own AI journey toward greater efficiency through smart automation.

Finnish/American company Sharper Shape — a DK&A client since 2019 — leads the way in developing AI-enabled software that generates digital twins for electric utility transmission and distribution systems.

The company recognized the potential for substantial efficiency gains by deploying powerful machine learning models as their first of many steps toward becoming an AI-powered organization early on.

At the outset, DK&A’s engineers and designers supported the team at Sharper Shape by replacing labor-intensive manually flown power line inspection with high-powered semi-automated drone flights. At the time, the drones still needed to be controlled by qualified pilots but were already capable of detecting defects autonomously and capturing images of the identified issues.

Seeing the immense potential of machine learning, Sharper Shape started to look for opportunities to unlock further efficiency gains through automation. In the process, the team identified two general areas for improvement in an attempt to enable fully autonomous drones:

  • Some of the images taken of defective insulators ended up being off-centered. In an attempt to correct this, pilots had to intervene manually.
  • Some planned routes for drone inspection flights significantly strained pilots due to demanding conditions. This also meant that pilots in the most unfavorable conditions were left to fend for themselves.

Repetitive tasks with a need for a high degree of accuracy paired with more or less readily available training data? — A perfect job to automate using machine learning.

How to make AI a reliable photographer

The most suitable input for the drones to determine the correct framing of shots was GPS coordinates. Or at least that was the initial assumption. Soon it became apparent that an accurate location determination was simply not feasible due to magnetic interference. And, to make matters worse, the measurement error differed from case to case, which made it hard to implement mathematical corrections.

Drone flying above the trees. There's a drone control showing power lines on the screen.

For human pilots, GPS coordinates are nothing more than a general reference point as they use their vision to decide whether a captured frame is usable. Using computer vision over location information seemed like a better approach to improving picture quality.

On a high level, the process at play is comparable to teaching a child. The machine learning model learns to distinguish insulators from other parts of an image through training. By feeding the model pictures with labeled insulators, the AI can learn to recognize them in even the most complex images.

Once the insulator was identified in the image, it was simple to rotate the camera gimbal to place it in the center of the frame. Machine learning enabled Sharper Shape to solve a problem universally, which had required manual changes on a case-by-case basis before.

Utilizing the power of AI

Drones are just one of several ways Sharper Shape has started to utilize Machine Learning and AI to increase efficiency. By developing an AI-aware business strategy, the company refined its identity and began to evolve toward becoming an AI-powered organization. Such a transformation would not have been possible without acknowledging the significance of the following:

  • having access to the right talent to drive innovation, and
  • the ability for advanced data management.

Sharper Shape has proven how effectively an organization can utilize the power of AI to further business goals. Succeeding in the two areas mentioned above enabled them to respond to the latest AI developments with tremendous enthusiasm and positioned them well to drive their AI development forward with confidence.

AI could be your future too

Embracing Artificial Intelligence and Machine Learning as long-term strategic factors has already become critical for any organization looking to remain competitive and innovative in an ever-evolving digital landscape. AI is not merely a fleeting technological trend or a one-time feature that can be implemented and forgotten about. AI might very likely prove an investment in an organization’s future success.

Envisioning how AI will transform your company, product, or industry should be an endeavor worth your time. Similarly to how we should have perceived the Internet when it emerged, we must adopt a holistic approach towards AI as a system of doing things to comprehend the magnitude of its potential impact fully.

Introducing and then scaling AI is a journey that involves executives and specialists, thinkers and engineers, business people, and customers all together. It is a journey that is open to organizations at all technology levels.

Unlock the value of AI

Learn more about our AI-powered products.

DK&A is a full-stack digital product agency guiding you from idea to impact.

By integrating cutting-edge technology with your products, such as generative AI like ChatGPT, we create exceptional, delightful experiences that give your brand a new edge and provide your customers with significant added value.

We can help you elevate your customer experience and achieve a lasting impact through ethical, customer-centric, and innovative approaches.