Exploring the Transformative Impact of AI in Cycling

Introduction to AI in Cycling

The days speculating about the potential arrival of artificial intelligence (AI) are gone. AI is here, already with significant transformative effects on our daily lives and AI in cycling is no different.

Most of us have probably seen the wide ranging, and sometimes extreme, coverage of AI technology. From the light-hearted pieces on how students are using Generative AI tools, like ChatGPT, to write essays, to the somewhat disturbing apocalyptic theories based on a certain movie with Arnold Schwarzenegger.

AI is going to have an impact on all aspects of our daily lives, beyond ‘the dog ate my homework’ theories or our version of Terminator 2: Judgment Day.

In cycling, where data has taken centre stage in recent years, AI is perfectly placed to drive the sport forward. In some cycling communities, AI is already deeply embedded, with people leveraging newly-created tools to improve performance in fine-tuning their kit.

Benefits of AI in Cycling

There are some quite straightforward benefits of AI in cycling, many of which are already been utilised.

• Route Optimisation: AI can analyse terrain, traffic patterns, weather etc. to suggest optimal routes for cycling based on the rider’s goals, preferred distance, elevation gain, training program etc. This allows cyclists to find the most effective routes for their specific requirements.
• Performance Tracking & Analysis: Sensors and AI can track detailed performance metrics like power output, cadence, heart rate etc and provide insights on how the cyclist can improve. AI can also detect weaknesses, or under performance, and give highly customised training and nutrition advice.
• Predictive Maintenance: By analysing vibration patterns and sounds from the bike, AI can identify parts that need maintenance or replacement before failures occur. This improves safety and riding comfort.
• Collision Avoidance: AI-powered cameras and sensors on bikes could detect potential collisions with cars or pedestrians and warn the cyclist or even automatically brake to prevent accidents.
• Group Co-ordination: For group rides, an AI system could help co-ordinate the group, provide guidance on optimal formations based on terrain, rider abilities etc. It could also alert riders about upcoming turns, stops and even traffic if there was an integrated traffic network.
• Enhanced Cycling Computers: AI assistants can provide verbal route alerts, weather updates, performance insights and other data to riders in real-time through integrated cycling computers. Imagine this on a Pro UCI race, say, the Tour de France.
• Autonomous Cycling: Though still in research stages, AI and autonomous technology could one day enable self-riding bikes that could follow routes or pelotons autonomously.

AI & Individual Configuration

One of the most important steps for any cyclist is choosing the right bike, and when it comes to the selection process, the frame is perhaps more important than the model or brand.

The problem is that bike fitters can sometimes be a little inconvenient, expensive, and sometimes just wrong. MyVeloFit is one of the companies trying to disrupt this market and help with choosing the right bike fit. This Canadian outfit used the expertise of Shopify’s ex-tech officer to help create a tool that gives users inch-perfect advice on fitting a bike. The AI in the system analyses videos over time, giving users constant advice on improvements.

In addition to assisting with frame configuration, it also extends to the kit that cyclists wear. To get optimal comfort and performance, cyclists can create an apparel design with AI.

Training with AI

AI is now also being used as a training tool for cyclists. AI Endurance is one of the companies pushing the envelope when it comes to generating individualised training plans for athletes.

Their tool creates a detailed trajectory for a cyclist in the run-up to a race, analysing the user’s exact requirements to create a minute-by-minute plan and reap the benefits of your hard work.

The AI technology responds to your performance, alongside that of all the other registered athletes, to further fine-tune your experience. The tool is constantly learning and adapting to your cycling, ensuring every training session is worth your time and effort.

These tools can also be used by anyone, the ‘regular Joe’ as our American cousins might say, ie. the person who is just like most of us, rather than a serious, almost semi-professional cyclist.

The community can sometimes be a little intimidating for new cyclists, which is why AI tools can be so useful to break stigmas and help you get going without adding excessive costs or any sort of embarrassment. With AI, users can build their perfect training schedule, whether it’s just to improve their health or work towards a more serious race.

Examples of Companies Using AI in Cycling

Artificial Intelligence isn’t new but its use in cycling probably is. However, despite this early stage technology being new to cycling, there are actually a fair number of companies already utilising AI in various cycling applications.

• Zwift: An online cycling training platform that uses AI for customised training plans and analysis of rider performance. Their Watopia platform uses AI bots as surrogate riders.
• Peloton: Their high-end exercise bikes use AI and sensor data to give riders personalised recommendations and adjustment of resistance levels for the best workout.
• Komoot: An online route planner that uses AI to analyse rider preferences and maps to suggest optimised routes for road, gravel or mountain biking.
• Sigma Sports: They make cycling computers like the Rox 12.0 that uses AI to offer on-device guidance, performance insights and predictive analytics.
• Tero Labs: Their AI-based bike-mounted sensors provide real-time feedback on posture and technique to improve cycling efficiency.
• HumanForest: An urban bike rental company using AI cameras for theft detection, rider identity verification and automated billing.
• Cobi: Their smart bikes integrate AI bike computers to provide anti-theft alerts, fitness data and even integrate with smart home systems.
• Amazon: They have patented autonomous AI bike couriers that can ride independently for deliveries requiring human hand-off at destination.

As you can see, AI in cycling is already in full-flow, and is being applied in cycling for route planning, performance tracking, predictive maintenance, training guidance and potentially even autonomous riding. We’ll likely see even more innovative applications as the technology evolves.

AI + Diet = Tour de France Victory?

Perhaps the most professional usage of AI in cycling today is how elite athletes are using it to modify their caloric intake. The Tour de France is a gruelling race, the most prestigious on the calendar, and one where millions of pounds, euros and dollars are at play.

When it comes to peak athletic performance, even the tiniest of margins will make the difference between victory and defeat.

When it comes to peak athletic performance, even the tiniest of margins will make the difference between victory and defeat. Kristian van Kuijk and colleagues, based at Maastricht University in the Netherlands, have recently introduced an AI-driven solution to ensure athletes get the perfect cocktail of calories for optimal cycling performance.

Using mathematical modelling and machine learning – a sub-division of AI – the team was able to help Tour de France cyclists put together a diet plan tailored specifically to them. Using the individual’s stats, and their race profiles, in addition to climatic conditions during a given stretch of the race, the tool can work out energy requirements along each stage.

Add a Touch of Dutch

… or so the tagline of a new AI-driven tool reads. Most of us are aware of the reputation the Dutch have: they cycle everywhere, and with a certain passion. The centre of Amsterdam is so full of bikes, that the city is opening up a new underwater parking area for cyclists.

The drive for more Dutch-style cities has been strong amongst cycling enthusiasts. Using a new AI tool, people can now see what that potentially looks like. The software, which calls for users to “add a touch of Dutch” to their streets, uses AI to populate an existing street plan with elements you’d find all over the Netherlands. It not only allows the individual to imagine what a cycling street would look like but also gives inspiration and guidance to city planners and political leaders alike.

Creating Safer Cities

Within the same theme of Dutch-style cities is the usage of AI in creating safer cycling in existing urban centres. In a recent hackathon, organised by the European Institute for Innovation and Technology’s Urban Mobility department, Cycle AI was presented as a potential solution for assessing road safety in cities.

Using artificial intelligence, the developers of Cycle AI can use images of roads to analyse cycling conditions and existing hazards, providing a safety score and recommendations for improvement.

Plenty of people are willing to cycle more, they just don’t feel safe when doing so. Existing traffic conditions are a big part of that. Using an AI tool to properly analyse existing conditions may be helpful for city planners and cycling infrastructure developers, assisting in the development of modern cities where the bike takes centre stage.

AI Cycling Products

There are a number of AI related products already available to cyclists and from some household names too, such as Garmin, Wahoo, Sigma and Tacx.

• Wahoo Elemnt Roam & Bolt: Cycling computers by Wahoo that use AI for performance analysis, route recommendation and real-time guidance.
• Sigma Rox 12.0: Advanced bike computer by Sigma with integrated AI for training, route planning and analysis.
• VeloComp PowerPod: Pedal-based power meter that uses AI for accurate real-time power tracking and customized feedback.
• Ride1Up AI Bike Computer: Budget cycling computer with AI for tracking metrics like speed, cadence, distance, route.
• See.Sense ICON2: AI-powered bike lights that collect riding data to guide cyclists on safer night routes.
• Tero Labs Tero: Sensor attaches to seat post to give real-time AI feedback on posture and pedalling efficiency.
• Hammerhead Karoo 2: AI-powered onboard cycling computer with route planning, analysis and training insights.
• Garmin Varia RTL515: AI-enabled bike radar that warns of approaching vehicles from behind.
• Bryton Rider S500: Cycling computer using AI to offer turn-by-turn navigation and training/nutrition advice.
• Cobi System: Integrates an AI bike computer with gear recognition and anti-theft alerts.
• Tacx NEO Bike Plus: Smart training bike utilising AI for immersive, realistic ride simulations and training.

So as you can see, there are quite a few products leveraging AI in various ways to enhance route planning, safety, training analysis and overall cycling experience. More innovation in this area is expected as the AI technology is advanced and adopted by cyclists.

Only the Beginning of AI Cycling

Of course, AI is still only in its infancy. Scary to some, incredibly exciting to others. The potential is clearly visible for cycling, and it will undoubtedly lead to improved individual and team performances.

But it goes beyond that. AI will change how both recreational and competitive cyclists approach the sport. It can make cycling more accessible, tailoring training programs for each individual, ultimately widening access. And that’s what we want, cycling expanding both as a sport and a way to get from A to B, improving our roads, and our health, and helping curb CO2 emissions.