Not as easy as the hype would have you believe.
For the first time, I used artificial intelligence to help plan an entire multi-day motorcycle road trip. Copilot is my go to AI tool but it could not create .gpx files so I had to use ChatGPT. Something to be aware of ChatGPT has some limitations…money limitations. Boy is it cash hungry.
Not just to answer, “What’s the fastest way from Point A to Point B?”
I wanted to see whether AI could actually help me plan a motorcycle trip the way I normally would: choosing the roads, spacing out fuel and rest stops, finding hotels, locating Harley-Davidson dealerships, adding interesting places to visit, avoiding roads I didn’t want to ride, and ultimately creating GPX route files for my Harley-Davidson navigation system.
The trip is from Ohio to Branson, Missouri, and back.
What started as a simple experiment with ChatGPT turned into a surprisingly detailed exercise in AI motorcycle trip planning.
It also taught me something important:
Using AI to plan a motorcycle road trip works much better as a conversation than as a single question.
How I Normally Plan a Motorcycle Road Trip
We have been taking long motorcycle trips for years, so trip planning isn’t exactly new territory.
Destination comes first. Then I start figuring out roads, daily mileage, hotels, fuel, food, interesting stops and all the other pieces that turn a line on a map into an actual motorcycle trip.
If you’re getting ready for a multi-day ride, my motorcycle road trip checklist covers many of the things worth considering before leaving home.
I’ve also previously written about the challenges of planning a long motorcycle ride. Apparently, nine years later, I’m still finding new ways to make motorcycle trip planning complicated.
This time I decided to see what AI could do with it.
The Rules I Gave AI for Planning the Motorcycle Trip
I wasn’t interested in simply finding the fastest route to Branson.
I wanted a route that made sense on a motorcycle.
Some of my requirements included:
- Breaks approximately every two to two-and-a-half hours
- A planned lunch stop every day
- Harley-Davidson dealerships close to the actual route
- Good overnight hotels, preferably Fairfield Inn or another Marriott property
- Reasonable riding days
- Interesting roads instead of automatically choosing the fastest interstate
- No unnecessary backtracking
- Specific roads or areas included when I wanted to ride them
- Specific roads and areas avoided when I didn’t
- GPX files that could be imported into the factory navigation system on my Harley-Davidson
If you’re planning your own extended ride, my tips for long motorcycle trips are also worth working into the planning process.
Then I started giving ChatGPT instructions.
And changing them.
And correcting things.
And adding more instructions.
AI Did NOT Get My Motorcycle Route Right the First Time
This may be the most important takeaway from the entire experiment.
Do not expect AI to create your perfect motorcycle route with one prompt.
We went through a bunch of different versions before arriving at the final routes.
At one point, AI wanted to route us through West Virginia.
I didn’t want to go through West Virginia on this particular trip.
New rule:
No West Virginia.
I wanted to use US Route 50 for the first portion of the outbound ride.
Another change.
That wasn’t an arbitrary choice, either. We’ve ridden Route 50 before, including during a previous motorcycle vacation on US Route 50, so I knew the type of riding I was asking AI to incorporate.
Then we started working on Harley-Davidson dealership stops.
Some dealerships didn’t fit the route.
Some needed to be removed.
One that initially appeared to be a possible stop had gone out of business.
Even the Overnight Hotel Changed the Motorcycle Route
Originally, our first overnight stop wasn’t quite where I wanted it.
I realized I wanted to get completely through Louisville on Day 1.
That way, instead of waking up the following morning and immediately dealing with Louisville traffic, we could start Day 2 already west of the city.
That changed our overnight stop to New Albany, Indiana.
It also demonstrated one of the things I began to like about using AI for motorcycle trip planning.
I didn’t have to start over. I just asked to Move the hotel west of Louisville and rebuild the route around it.
The AI took the new requirement and adjusted the plan.
That’s when this started feeling less like a search engine and more like having another person sitting at the table helping me work through the map.
Planning a Different Motorcycle Route Home
Once the outbound trip was mostly finished, I decided I didn’t want to ride the same roads home.
So I gave ChatGPT another instruction:
Don’t backtrack over roads we’ve already ridden unless necessary.
That turned the return trip into its own three-day motorcycle tour.
And that is where things became much more interesting.
Instead of simply trying to get home, we started adding places worth riding to.
The return route eventually included:
- Metropolis, Illinois, and the giant Superman statue
- Towers’ Bourbon Barrel Harley-Davidson in Paducah, Kentucky
- The Nicky Hayden Memorial in Owensboro, Kentucky
- Man O’ War Harley-Davidson in Lexington
- Kentucky’s Red River Gorge
- Nada Tunnel
- KY-715 through the Gorge
- The Ohio River Scenic Byway
Now we weren’t just using AI to calculate directions.
We were using it to design a motorcycle road trip.
Using AI to Route a Motorcycle Through Red River Gorge
Red River Gorge became one of the best examples of why detailed instructions matter. I wanted the motorcycle route to go:
KY-77 → Nada Tunnel → KY-715 → KY-15
I specifically did not want to ride through Nada Tunnel, turn around somewhere and come back through the tunnel again which Chatgpt suggested initially.
The goal was a continuous ride through the Gorge.
That meant the route needed more than a start and destination.
It needed enough intermediate points to tell the navigation system exactly what I was trying to accomplish.
That issue became even more important when we moved from Google Maps to the Harley-Davidson GPS files.
AI Checked Harley-Davidson Dealership Locations and Hours
I originally asked for Harley-Davidson dealerships within approximately five miles of our motorcycle route.
Finding the dealership was only part of the job.
Would it actually be open when we got there?
That question eliminated some potential stops.
During the planning process, we began checking operating days and hours and comparing them with our expected travel days.
That’s an important lesson for anyone experimenting with AI travel planning:
Never assume a business suggested by AI is still operating or will be open when you arrive.
The same applies to hotels, restaurants, attractions and even roads.
Verify important information before the trip.
Creating GPX Files for My Harley-Davidson Boom! Box Navigation
This may have been the most interesting part of the entire project.
Once the routes were established, I asked ChatGPT to create GPX motorcycle route files that I could use with the factory navigation system on my Harley-Davidson.
A GPX file can contain waypoints, routes and other geographic information that a GPS/navigation system can interpret.
But creating a GPX file is only half the battle.
Motorcycle GPS users already know the problem:
Your GPS sometimes thinks it knows where you want to go better than you do.
You can carefully plan a great motorcycle road and then watch the navigation system calculate a completely different route between two points.
So we went another step.
Using GPX Shaping Points to Control the Harley Route
We added additional GPX shaping points to encourage the Harley-Davidson navigation system to follow the roads I actually selected.
Three sections received special attention.
Southern Missouri
I didn’t want the return trip simply putting us back onto the roads we used going to Branson.
Additional points helped define the different return corridor.
Ohio Route 7
Our final day follows OH-7 along the Ohio River.
Because there are bridges crossing into West Virginia, a GPS could potentially decide that crossing the river and returning later would save time.
But one of the rules for this trip was to stay out of West Virginia.
So the GPX received additional shaping points along OH-7 to encourage the Harley navigation system to remain on the Ohio side of the river.
That was something I hadn’t expected when this experiment began.
I wasn’t simply asking AI:
“Where should I ride?”
I was asking it:
“How do we build the navigation file so my motorcycle understands where I want to ride?”
That’s a much more interesting use of AI.
We Even Created Motorcycle Tank-Bag Route Cards
I like technology.
I also don’t completely trust technology.
Those two ideas can coexist.
So, in addition to the GPX files, we created a one-page motorcycle tank-bag route card for every riding day. (see bottom for one of the tank-bag cards)
Each card contains the important information I might want at a glance:
- Starting point
- Destination and hotel
- Major roads
- Fuel and rest stops
- Lunch stop
- Harley-Davidson dealership stops
- Points of interest
- Important routing notes
If the GPS has a bad day—or I simply want to see what’s coming next—I still have a printed reference.
New technology meets an old-fashioned piece of paper.
Sometimes that’s the best combination.
What AI Did Well When Planning a Motorcycle Trip
After all these iterations, I found several things AI did particularly well.
The biggest was keeping track of an increasingly complicated set of requirements.
I could tell it:
Don’t go through West Virginia.
Use US-50.
Move the hotel.
Remove that dealership.
Find another Harley dealer.
Put lunch here.
Add a fuel stop.
Don’t backtrack.
Add this attraction.
Go through Nada Tunnel.
Stay on KY-715.
Keep the return route different.
Build the GPX file.
Add more shaping points.
Create a tank-bag card.
And the motorcycle trip continued evolving.
That’s where AI provided the most value.
It didn’t magically discover the world’s greatest motorcycle route.
It helped manage dozens of small motorcycle trip-planning decisions until they became one complete itinerary.
Where AI Motorcycle Trip Planning Can Go Wrong
There is another side to this experiment.
I would never recommend blindly following an AI-generated motorcycle route.
AI can make mistakes.
In fact, I caught several during this project.
Information can also become outdated.
A dealership can close.
A restaurant can change its hours.
Construction can close a road.
Weather can make a fantastic motorcycle road a terrible idea.
And a GPS can interpret a GPX file differently than expected.
I’ve seen firsthand how quickly road conditions can change. My experience riding Ohio Route 555 when chipseal made a great motorcycle road hazardous is a good reminder that the route you planned months ago may not be the road you find when you arrive.
Before leaving on this trip, I’ll also give the motorcycle the normal inspection. My motorcycle tire safety checklist is one part of that pre-trip routine.
Most importantly, I’ll load each GPX route into the Harley-Davidson navigation system and preview it.
If the Boom! Box decides to take us somewhere we didn’t plan to go; I’ll will not be able adjust the GPX. Some of these routes are long and I will likely miss errors that I cannot adjust for on the road due to the limitations of GPS systems.
AI is a tool for planning the ride. The rider is still responsible for the ride.
Is AI Worth Using to Plan a Motorcycle Road Trip?
Based on the planning stage of this experiment, maybe.
But not for the reason I expected.
AI didn’t eliminate the work involved in planning a motorcycle trip.
It changed how I did the work.
Normally I might have Google Maps open along with hotel websites, dealership searches, restaurant searches and GPS route-planning software.
Meanwhile, I’m trying to remember why I chose one road three days earlier.
This time, much of that process became one continuing conversation.
Move this. Add that. Don’t use this road. Find something near this route. Move tonight’s recommended hotel. Add somewhere interesting to stop. Make tomorrow shorter. Now make the route work with the Harley GPS.
That conversational approach turned out to be the real advantage.
The Real Test: Riding an AI-Planned Motorcycle Trip
Right now, everything looks good on the computer.
We have:
- Multi-day outbound and return motorcycle routes
- Hotels
- Fuel and rest stops
- Lunch stops
- Harley-Davidson dealerships
- Points of interest
- Google Maps routes
- GPX files customized for the Harley-Davidson navigation system
- Printed tank-bag route cards
But motorcycle trips aren’t ridden on computer screens. They’re ridden on pavement.
The real test comes when we actually ride the AI-planned motorcycle trip.
Did the riding days feel right?
Were the breaks spaced properly?
Were the hotel choices good?
Did the Harley navigation system follow the GPX routes?
Were the shaping points enough to keep us on the roads we selected?
And, most importantly:
Finding out whether AI can actually help create a great motorcycle road trip is something else entirely.
Stay tuned.
Ride on, Ride Safe
Leave a Reply