Bike Computer & Navigation: Plan and Ride Routes

A bike computer is not an accessory — it is the cockpit of your training. Even so, most riders choose one on the wrong criterion: screen size. The things that actually matter later appear in no product listing — whether the unit survives a five-hour ride without recharging, whether navigation still works after one wrong turn, and whether you end up with access to your own data. This article sorts the features by their value for structured training and shows what you really need to start.

The six core functions — everything else is decoration

A bike computer for ambitious training has to do exactly six things reliably. Anything beyond that is comfort, anything missing is a deal-breaker.

• Reliable recording. GPS recording with at least two satellite systems (GPS plus Galileo or GLONASS) and an altimeter. A barometric altimeter is far more accurate than GPS elevation and therefore the basis of every climbing metric — on a gran fondo with 2,000 metres of climbing, the errors otherwise reach into the three digits.
• ANT+ and Bluetooth at the same time. No device should pin you to one sensor standard. Anyone running a chest strap from one brand and a power meter from another wants to pair both — and to be able to carry a spare sensor.
• Battery life of 15 hours and up. Manufacturer figures assume frugal settings. With the map open, active navigation, backlight and paired sensors, a unit typically loses 30 to 50 percent of its runtime. For a six-hour gran fondo plus the ride to the start and home you need a real 15-hour buffer — otherwise your training depends on a power bank.
• A display you can read in full sun. A transflective display is more readable in direct sunlight than an OLED because it uses reflected ambient light instead of fighting it — and it uses less power doing so. If you ride in the sun a lot and plan long days, transflective wins.
• Offline maps. Map data stored on the device instead of streamed over mobile data. In the mountains you lose signal exactly when you need the map most.
• Open data export. Your ride data has to be exportable as a FIT file without a subscription. Every device does this — the platform behind it is where they differ.

Navigation: the difference between an arrow and understanding

There are two fundamentally different approaches to navigation, and product comparisons rarely separate them properly. The simple version is breadcrumb navigation: you follow a loaded route and the device warns you when you drift off course. That works surprisingly well as long as you never go wrong — and surprisingly badly the moment a road is closed, you take a detour, or you deliberately choose a different junction. Because the device knows exactly one route: yours.

The second version is real map navigation with recalculation: you see the map, you can deviate, and the unit guides you back onto the route. That costs more processing power, more battery and more money — but it is the decisive difference between a training device that lets you ride freely in unfamiliar terrain and one that chains you to a single line. For training camps, gran fondos with a changed course, or simply riding by feel, recalculation is worth the extra cost.

A third point that gets lost in reviews: turn prompts in the right order. A good unit warns early, then warns again just before the junction, and displays the distance to the next action larger than the remaining ride distance. If the screen occupies you with total kilometres 200 metres before a roundabout, the interface is wrong — not the rider.

Route planning: purpose first, line second

The most common mistake in route planning is the order. Most riders build a beautiful route and then look at what kind of session comes out of it. It is the other way round: first decide what the session has to deliver, then build the route around it.

• Base session (zone 2, 2 to 4 hours): Flat to gently rolling terrain, few junctions, ideally uninterrupted. The goal is constant load without stops — a stop every ten minutes destroys the training stimulus. Planning rule of thumb: 40 to 60 metres of climbing per 10 kilometres in rolling terrain makes for a pleasant easy ride; above 100 metres per 10 kilometres it becomes hill training by accident.
• Threshold or interval session: An even route with long straight sections and no crossings. A 20-minute block interrupted three times by a village centre is not a 20-minute block. Plan intervals on a loop you can ride several times — then you are not dependent on one perfect stretch.
• Hill training: Choose the climbs first, then plan the approach. An 8-kilometre climb at 6 percent is a different training stimulus from four climbs of 2 kilometres at 9 percent, even when total elevation is identical. For longer threshold work you need even, moderately steep climbs; for strength endurance, short steep ramps.
• Gran fondo preparation: Break the target course into sections and get to know each one. Anyone who has ridden the decisive 60 kilometres twice beforehand plans fuelling and pacing completely differently from someone who only knows the elevation profile.

For planning itself: 25 to 32 km/h is a realistic average range for a trained recreational rider on flat ground — headwind, climbing or group rides shift that considerably. Plan breaks every 60 to 90 minutes and check the route for roadworks, closed sections and surface before you ride. Good route planning follows the same logic as running: a clean loop beats hitting an exact kilometre target.

Sensors: what actually changes anything

Bike computers are usually sold in bundles with sensors, and this is where the most money gets spent wrongly. The order by value is clear:

• Heart rate (chest strap) — essential. A chest strap measures far more reliably than an optical wrist sensor, especially during intervals, in the cold, and when your hands are on the bars. Heart rate is the basis of every zone-based session — and therefore of every structured workout. See heart rate zones.
• Cadence — very worthwhile. A cadence sensor is cheap and delivers one of the most useful feedback signals there is. Most recreational riders pedal at 70 to 80 rpm, while 85 to 95 rpm is more economical in most situations. If you do not know your cadence, you cannot change it. Details in cycling cadence.
• Power meter — the biggest step. A power meter is the most expensive single purchase, but it changes training more than any other device: load becomes objective, independent of wind, elevation, daily form and cardiac drift. Intervals can be dosed precisely and training stress captured exactly. For anyone training to a structured plan, a power meter is the most sensible investment after the bike itself — see FTP testing and watt training.
• Speed sensor — dispensable. Redundant on a GPS bike computer. Only relevant if you spend a lot of time on the indoor trainer or ride tunnels without reception.
• A second power meter — pointless. Two sources for the same quantity only create the question of which value is right. One unit per bike, with the option to swap it, is the cheaper and cleaner solution.

TSS and training load: from the end of the ride to actual steering

A bike computer records, but it does not steer. The real work starts afterwards: turning raw data into a load number and placing it in its context over time. In cycling this is especially clean via power, because watts are a direct physical quantity.

One hour at threshold equals 100 TSS by definition. From that come reference values: an easy 90-minute ride sits at 50 to 70 TSS, a three-hour base ride at 120 to 160, a hard 90-minute interval session at 120 to 150. Collect those numbers over weeks and two curves emerge: the short-term build-up of fatigue (ATL) and longer-term fitness (CTL). The difference — form (TSB) — tells you when a training block is productive and when you are overreaching. Details in TSS, CTL, ATL and TSB.

The practical value: without this framing, every ride ends with a feeling instead of a number. With it, you know before the session how much load it should carry — and afterwards you can check whether it did. That the ride lasted 4:12 hours says nothing about its effect; that it cost 165 TSS says everything.

Data protection: who sees your ride data

This is where the device discussion ends and the platform question begins. A bike computer stores locally — but as soon as you sync the ride, position, heart rate, power and timestamps travel to a server. From that data a movement profile can be reconstructed: where you live, when you train, when you travel. That is more sensitive than most people assume.

• Server location. EU servers mean GDPR applies and you have a point of contact in Europe. With providers hosting outside the EU, the data transfer becomes your own legal problem — details in GDPR and fitness apps.
• Export instead of lock-in. Your FIT files have to be fully exportable. Anyone who cannot take their training history with them is tied to a platform, even if that is written nowhere.
• Genuine deletion. A delete-account button that really removes the data and does not simply hide the profile. That is mandatory in the EU, but not implemented equally everywhere in practice.
• Share with care. Live tracking links and segment comparisons are handy, but they make your position public. If you use live tracking, use it with a delay or share it with a limited circle only.

The five most common mistakes when starting out

• Buying by the display. A big, bright screen is worthless if the battery is flat after eight hours and navigation fails without signal. Decide on runtime, sensor compatibility and maps — the display comes last.
• Showing every data field. A cockpit with twelve fields does not get read in a race. Two to four values per page are enough, and only the most important one should be large. During an interval that is the target value, the current deviation and the time — not the elevation total.
• Going long without a spare battery. A six-hour gran fondo plus the ride to the start and home breaks every specification. A small 5,000 to 10,000 mAh power bank weighs under 200 grams and belongs in the saddle bag on long days.
• Checking the route on the bike. What looks like a quiet country lane on the map is sometimes a gravel track at 14 percent. Look at street view or satellite imagery beforehand, and ride new routes the first time without any time pressure.
• Recording without reviewing. The most expensive mistake, because it costs nothing and therefore never stands out. A bike computer whose data never feeds into a load and form curve is an expensive speedometer.

How Peakora works with this

Peakora does not replace a bike computer — it is the layer above it. You record with your device as before, and the ride is placed into load management: duration and intensity become a TSS value, and over the weeks that builds the form curve from fitness (CTL), fatigue (ATL) and form (TSB). That way you can see before a long ride how well it fits your current block, and after a hard week whether the next session should be easy or intense.

Because the plan is computed deterministically rather than guessed, the load stays traceable: for every session you can see why it was dosed the way it was. That applies equally to running, cycling and swimming — one shared load curve across all disciplines instead of three separate numbers. Training data sits on EU servers, can be exported and deleted at any time. For more on the underlying system, the Peakora glossary covers the terms, and for preparing specifically for a long race the gran fondo training plan is the place to start.

Frequently asked questions about bike computers and navigation