Safety at climbing walls depends on a simple action that many people overlook: attaching to the autobelay before starting to climb. Lizcore has developed a device that combines cameras, machine vision, sensors, and NFC identification to ensure that the climber is securely attached and to alert when someone climbs without being properly secured. In chapter 82 of “Toque de Ingenio,” Edgar Casanovas explains how he identified the problem after 25 years as a guide and instructor, how he progressed from a shoebox with an Arduino to a product installed in several countries, and what he would do differently today.
Guest: Edgar Casanovas, who is introduced in the episode as the founder of Lizcore. Interview published: July 8, 2026. Episode: 82. Duration: 1 h 7 min.
In this episode:
- Why forgetting, rather than recklessness, is the cause of accidents in indoor climbing.
- How a camera checks the harness, carabiner, and anchor before allowing someone to climb.
- What did Lizcore learn about patents, minimum viable product, and market entry?
Edgar Casanovas brings a device for self-monitoring to the interview and explains its functionality during each check.
Safety in indoor climbing: the accident that begins with a lapse in concentration.
Edgar Casanovas began climbing at the age of 12 and has dedicated 25 years to the sport as a guide and instructor. He states that indoor climbing has grown rapidly: there are now 465 climbing walls in Spain, and the centres are struggling to keep up with the increasing number of new users.
Working at a climbing wall, he observed that users were coming in large numbers, but they weren’t staying, because there was no way to see their progress or compete. And a series of accidents began to recur, always following the same pattern.
Someone removed the safety harness from the autobelay to speak with a friend. They looked at the clock, returned to the wall, and ascended without having secured themselves. Upon arriving at the top, they were convinced that the rope was holding them, insisting that this was not recklessness, but simply a lapse in memory, comparing it to the seatbelt warning in a car.
The autobelay is the device that has enabled climbing without a partner – it gathers the rope as you ascend and instantly stops a fall. Its convenience has increased its use, and consequently, the number of accidents. The solution couldn’t be an app, as nobody climbs with their mobile phone in their hand.
Validate the idea outside of your circle of friends.
Edgar shared the idea with the parent of one of his students, who turned out to be an engineer. Within about two weeks, he had an Arduino with lights inside a shoe box. He checked that there wasn’t anything similar and, with the help of a designer and a computer expert, he built the first prototype, which was launched in 2023.
When he showed it to the breeders, everyone told him it was very beautiful. His wife pointed out the bias: it was like asking a friend if his daughter was pretty. The managers were colleagues he had climbed with throughout his life, and their opinion didn’t validate anything.
He decided to present it at an international fair in Germany, where nobody knew him. Many visitors didn’t believe that technology had a place in that sport. But when he asked if this was the future of climbing, the answer was yes. The following year, he returned with a functional prototype.
The team filed a patent and an international PCT application. When questioned about this decision today, Edgar is clear: he would not repeat this process. He believes the money could have been used to develop and launch the product more quickly, and, in his opinion, speed to market is what truly makes the difference.
machine vision to check that the climber is securely tied in.
The Safety Gate device, developed by Lizcore, is installed where the carabiner hangs. Since the autobelay always pulls upwards, the device holds the strap in place, freeing the carabiner so that the user can attach it.
Then, a camera comes into play. machine vision. According to Edgar, the system first checks for the presence of a human with a harness, then locates the ventral ring – the only one suitable for attaching – and finally verifies that the carabiner is in contact with it and locked. Only then does a latch release the strap.
The process doesn’t end there. The camera needs to verify that the carabiner pulls upwards with approximately 2.5 kilos of force, and only then does the device turn green. This all happens in a matter of seconds. Detecting the presence of a person also works in reverse: if someone is below the climber descending, the device will sound an alarm to prevent them from approaching.
In the event of a lapse, there’s a second, upper camera monitoring the wall. If a foot is detected in motion, and the device hasn’t delivered a strap to a bound person, a visual alarm is triggered. This is then repeated to nearby devices, the volume increases, and a notification with the location is sent to the mobile phones of the monitors. This alert arrives before the climber reaches eight meters.
Sensors that transform usage into maintenance.
In addition to monitoring the climber, the device also monitors the autobelay itself. It counts the number of times the tape goes up and down, allowing for maintenance to be scheduled based on usage rather than a fixed calendar. Edgar notes that an autobelay in Madrid experiences significantly more use than one in a rural area.
A built-in scale measures the force with which the autobelay pulls on the tape. If it suddenly pulls with more or less force than usual, the system alerts that it needs repair. And a device in poor condition can be remotely locked to prevent the release of the tape, even if a secure carabiner is in place.
This combination of camera, lock, and sensors within a single system is an example of Electronic engineering Applied to a specific environment. The usual alternative is a tarp hung on a carabiner: this requires removing it to climb, but no one checks to ensure the user has secured it afterward.
Identify the climber with NFC and gamify the progression.
Lizcore’s second product, the Grader Star, was designed for use within a specific area and connects via cable to the Safety Gate. Users are identified using an NFC-enabled wristband, hat, or t-shirt. Once authorized – after completing the initial training – the system initiates the checks.
With that identification, every climb is recorded in the app: completed routes, comparisons with other users, and challenges from brands. You don’t need to touch anything while climbing, as a scanner detects when hands are on the last dam in 3D.
Edgar explains the business logic. The cybersecurity market is limited, because resources are finite; gamification allows us to create endless games and challenges. Therefore, the Grader Star was initially designed, already with a connector for a future security system.
The market, however, wanted the opposite. Grader Star launched in 2024 and sold poorly. Safety Gate launched in 2025 and is gaining ground, because managers prioritise products that prevent harm over those that retain customers.
Training artificial intelligence and convincing a traditional sector.
According to Edgar, the most challenging technical aspect has been teaching the camera to recognize the correct patterns. It all began by photographing a group of friends with and without harnesses. Today, the image library continues to grow, and the system is now faster than simply removing a sheet and putting on a harness.
Another challenge lies in the cultural aspect. Experienced climbers feel that the equipment is treating them as if they are incompetent, and they respond by saying that they always tie in. Edgar recalls that a friend of his also always tied in, until the day he didn’t. He distinguishes between the mountain, where he accepts the risk, and the climbing wall.
“The risk in an indoor sport taking place in the middle of a city must be zero”.
Edgar Casanovas 55:53.
When asked about his biggest mistake, Edgar responded that it was the slowness. His desire to make everything beautiful and to achieve hardware, software, and an application simultaneously led to a delayed validation process.
“To do one thing, just one, extract it, but don’t sell it, but demonstrate that it’s feasible”.
Edgar Casanovas 65:58.
What can another development team learn?
The Lizcore case reveals four key lessons applicable to other products utilizing sensors and machine vision:
- Test with strangers. People familiar with the industry praise any prototype, and a fair where you don’t know anyone provides a more reliable indicator.
- Design for the absent-minded user. A chain of checks and escalating alerts is more effective than assuming that the user will act correctly.
- Remove one thing first. Demonstrating a product’s viability before developing the next one shortens the path to the first sale.
- Convert this usage into maintenance. Tracking cycles and measuring strength allows you to assess a piece of equipment when it actually needs it, not just when the calendar dictates.
At the time of the interview, Lizcore was sending Safety Gates to four airfields in Italy, installing them in Mexico and Barcelona, and negotiating with centres in Madrid, Holland, and the United Kingdom. They were present in seven countries and preparing a round of investment for manufacturing and marketing.
In i-mas, we develop. products which incorporate electronics, machine vision, mechanics, and software. If you are working on a device that needs to detect, measure, and alert in a real-world environment Tell us about your project..
To continue reading: Sensor and application for measuring force based on speed, the case of Vitruve..
Source of article: Interview with Edgar Guerrero by Edgar Casanovas on Toque de Ingenio. The excerpts cited link to the minute of the conversation. The recommendations for other projects are a summarised editorial from i-mas.
