#15: Footwear's AI Healing Touch, Deepening Virtual Realms, and Scooting Ahead with AI Voltages
Hey there, AI friends! 🌅
Welcome to Issue #15 of Good AI Vibes. We've got more cool AI stories for your business. Keep reading and find new ways to use AI.
Here's what we have this time: 👇
👟 Healing in Every Step: The AI Revolution in Footwear
🥽 Depth Matters: AI's Breakthrough in Augmented Reality
🛴 Voltage and AI: The New Tune for Electric Scooters
Know someone who'd like this? 💌
Sharing Good AI Vibes is quick. It helps us and gives others new ideas for their business.
Ready for some great AI stories? Let's get started!
Healing in Every Step: The AI Revolution in Footwear
What's the critical hurdle?
Meet Martina, a dedicated manager at a health institution that specializes in rehabilitating patients with lower limb disorders. She often meets patients grappling with conditions like osteoarthritis, affecting their knee, hip, or ankle. These disorders not only restrict their physical movement but also dampen their quality of life. Martina's institution is always on the lookout for cutting-edge diagnostic methods, which can ensure the best possible care for their patients. One of the ongoing challenges she faces is swiftly and accurately categorizing these disorders so they can customize treatments accordingly. Every misdiagnosis or delay means a patient might not get the best possible care, which could result in prolonged recovery or inefficient treatments.
How was this initially tackled?
The initial approach was using traditional medical examinations and monitoring methods. Patients would be subjected to a series of manual tests, observation of their walk, and feedback about their pain. It was essential but time-consuming. Doctors would physically assess joint flexibility, pain points, and walking patterns.
Imagine a patient being asked to walk back and forth several times while a team of specialists observes her gait, takes notes, and then deliberates on the findings. It's efficient to some degree but lacks the finesse of precision.
Why did the initial approach fall short?
Though the traditional methods have been trusted for years, they come with a set of challenges. Observational methods can sometimes miss subtle nuances in a patient's walk or stance, leading to potential misdiagnosis. Personal biases and differences in interpretation among doctors could also play a role. This approach also took longer, causing delays in devising the best treatment plans.
Let's say two specialists observe a patient’s walk. One might focus on her knee movement, while the other might be more concerned about her foot placement. Both are crucial, but the varied focus could lead to differing opinions on her condition.
How did AI revolutionize the solution?
Enter the game-changer: AI-powered footwear embedded with special sensors. These aren't just any shoes; they are intelligent diagnostic tools! When patients walk wearing these shoes, the sensors capture the pressure and nuances of their gait. The data is then processed using AI, which meticulously analyzes every step and stride. And guess what? It doesn't just say there's a problem; it precisely categorizes the disorder, whether it's in the knee, hip, or ankle.
Imagine a patient wearing these sensor-equipped shoes and taking a simple walk. By the time she's done, the AI already knows the exact nature of her lower limb disorder. It's like having a team of specialists in the sole of a shoe!
With a whopping accuracy rate of 96% (better than many traditional methods), this approach ensures that patients get diagnosed faster and more precisely. It's a win-win: Health institutions can provide better care, and patients can step into a healthier future more confidently.
Main reference: Siddiqui, H. U. R., Nawaz, S., Saeed, M. N., Saleem, A. A., Raza, M. A., Raza, A., Aslam, M. A., Dudley, S., et al. (2024). Footwear-integrated force sensing resistor sensors: A machine learning approach for categorizing lower limb disorders. Engineering Applications of Artificial Intelligence, 127(Part A), 107205.
See other references at the end.
Depth Matters: AI's Breakthrough in Augmented Reality
What's the critical hurdle?
Meet Bianca, a project manager for a leading firm that develops augmented reality (AR) tools for interactive tourism. Her team creates apps where users can point their smartphones at historic sites and see AR pop-ups, offering detailed explanations about the site's significance. The core challenge? Creating a true 3D feel where objects closer to the viewer look nearer and those further away appear distant.
Imagine someone visiting the ruins of an ancient amphitheater. With a typical AR app, a digital gladiator might appear flat, as if he's merely a sticker on the scene, rather than part of the real, 3D world. This breaks the immersive experience and feels less authentic for the user. Moreover, this misinterpretation of depth can be bad for business. In a market where immersion and realism are critical, having a 2D-feeling AR can cause Bianca's firm to lose potential users to competitors, affecting the bottom line.
How was this initially tackled?
Before diving into the advanced world of AI, Bianca's team used stereo imaging. This involves using two images (think of it like our two eyes) to gauge the depth and recreate a 3D scene. For instance, to display that digital gladiator, they'd capture two slightly different angles of the amphitheater and use software to stitch them together, creating a sense of depth.
Why did the initial approach fall short?
While stereo imaging sounds effective on paper, it had its drawbacks. For one, it requires the capturing of two separate images – a logistical challenge when building a large-scale AR database. Furthermore, scenes change: trees grow, buildings get renovated, and monuments undergo repairs. This means that the stereo images need constant updating, a time-consuming and costly affair. Despite their best efforts with stereo imaging, the team still faced a hurdle. The AR experience was not as seamless and realistic as they desired. It often misinterpreted distances, causing some AR objects to float oddly or appear out of place.
How did AI revolutionize the solution?
Enter AI. Instead of relying on two images, AI could estimate depth using just one image. By leveraging vast datasets derived from video games (a novel and ingenious approach!), the AI learns to perceive depth in a single image, much like how humans infer depth even when closing one eye.
For Bianca's AR app, this means when a user points their smartphone at the amphitheater, the AI quickly interprets the depth of the scene in real-time. This creates a more immersive AR experience, with our digital gladiator appearing as if he's genuinely standing in the amphitheater, rather than just being superimposed onto it.
This AI approach not only elevates the user experience but also offers a cost-effective solution. Without the need for dual imaging, Bianca's team can scale their AR experiences faster and ensure that they remain top-tier in the competitive AR market.
Main reference: Haji-Esmaeili, M. M., Montazer, G., Montazeri, A. I., et al. (2024). Large-scale Monocular Depth Estimation in the Wild. Engineering Applications of Artificial Intelligence, 127(Part A), 107189.
See other references at the end.
Voltage and AI: The New Tune for Electric Scooters
What's the critical hurdle?
Meet Emilio, the operations manager for a bustling electric scooter-sharing service in a metropolitan city. Ensuring the health and longevity of the lithium-ion batteries in his fleet of scooters is one of his primary concerns. The SOH (State of Health) of these batteries is paramount, as it determines how long a scooter can run between charges and its overall lifespan. A deteriorated battery means a scooter might not make it to its destination, leaving riders stranded and frustrated. Imagine a daily commuter, who relies on Emilio's scooters for her last-mile commute. One day, she's late to a meeting because the scooter's battery drained faster than expected. Such instances can tarnish the company's reputation and hurt business.
How was this initially tackled?
Before AI came into the picture, Emilio depended on the scooter's basic battery management systems (BMS) that gauged measurements like current, voltage, and temperature to gauge battery health. It was akin to gauging the health of a garden by just looking at it from a distance, without checking the soil or the health of individual plants. If a scooter was used for a particular number of rides or hours, the BMS would predict its battery's health based on that usage and the temperatures it experienced.
Why did the initial approach fall short?
While useful as a general gauge, these traditional systems often lacked precision. Emilio observed that many scooters, which the BMS marked as having healthy batteries, actually had batteries that were inefficient or close to their end-of-life. This led to unexpected downtimes, unplanned battery replacements, and, worst of all, disgruntled users.
How did AI revolutionize the solution?
Unlike the conventional method, which provided a generalized estimation not accounting for the nuanced and intricate variances that each battery might experience, the AI system provides a detailed, tailored health report for each individual battery. It's a real-time solution, which means that as soon as there are deviations in the battery's performance, the AI picks it up, allowing for immediate action. This level of precision wasn't achievable before. By processing and analyzing vast amounts of data in real-time, AI offers an unparalleled understanding of battery health, ensuring longer life, more reliable performance, and fewer unexpected downtimes for the scooters. This, in turn, translates to more satisfied riders and efficient operations for the business.
Main reference: Mazzi, Y., Ben Sassi, H., Errahimi, F., Andre, D., et al. (2024). Lithium-ion battery state of health estimation using a hybrid model based on a convolutional neural network and bidirectional gated recurrent unit. Engineering Applications of Artificial Intelligence, 127(Part A), 107199.
See other references at the end.
Good AI Vibes Cartoons 😅
In case you missed our last edition, catch up on all the insights from Good AI Vibes #14 right here!
📌 Just a quick reminder: Share Good AI Vibes with your community and stand a chance to earn FREE premium subscriptions 🎁 with our Referral Program. Curious about how it all works? 👇