Built for cars. Banned for bikes.

Stop 01 · The question

Why the scooter stays parked

Transportation is the largest single source of greenhouse gas emissions in the United States, at roughly 28 percent of the total, and more than half of American trips are under three miles, the exact range a bike or scooter covers best. The climate math is settled. The campus math is not. On many university campuses micromobility use stays low, and some students who could ride still take a car to class.

Ask why and the first answers are the obvious ones: theft and safety. Both are real. A U-lock that holds a bike overnight does little for a scooter someone can simply carry off, and most streets still ask a 15 mph rider to share space with 40 mph traffic.

But theft and safety are not where most rides die. They die earlier, over things that never show up in a crime statistic: there was nowhere to park it, nowhere to charge it, and the fastest route was closed to riders before 9 a.m. The rider was rarely the problem. The campus was.

Stop 02 · The real barriers

The barriers that don't make the news

We rank the barriers students actually hit, drawn from what riders and non-riders tell us rather than what we assume. Three come up before any of the headline ones.

The first mile · Before the gate

The map ends at the property line

Most riders do not start on campus. 17 of our 31 respondents commute in, so the trip begins on a city street, and California has already graded that street. The Streets and Highways Code sorts every bikeway a city builds into four classes, from a path fully separated from traffic down to a road with a stencil painted on it. A commuter can look up which class their route carries before they leave the house.

Another SYN 100 team spent the quarter on exactly that leg. Bridging the Gap, by Anthony Tong, Alan Wang, Ari Pasha and Owen Molloy, mapped the neighborhoods around UC San Diego, counted traffic at six locations, and rated six ways onto campus. Their work answers the question ours never asks: which gate should you use. Set their findings beside ours and a third thing appears that neither project sees on its own.

Every row above is their finding, not ours. We cite it because the first mile is the half of the trip our own fieldwork never covered.

At five of the six locations they counted, about nine in ten of everything moving was a car. The one exception is Athena Way, their quietest crossing, where a little over half was car traffic and 41% was people on foot. It is the only gate that behaves like a footpath instead of a road, and it is the one nobody drives to.

A guide can tell you which class of lane carries you in. It cannot tell you how much of that network is worth having. So we went and measured it. Every figure in the next two paragraphs is ours, computed from SANDAG's own published route layers, and the arithmetic is re-run from the raw lengths before every release.

San Diego County has 1,976 miles of mapped bike route. 188 of them, fewer than one mile in ten, are a path set away from the roadway. The remaining 90% puts a rider in the road with cars, beside them, or on a recommendation. Whatever is happening to a student on that commute, it is happening within a few feet of moving traffic for nine tenths of the network.

Then there is the network the region has already agreed to build. SANDAG's Adopted Regional Bike Network runs to 556 miles. In the agency's own layer, 214 miles are marked completed and 342 miles are still marked future. 62% of the bike network San Diego has formally adopted does not exist. Not rejected, not under debate, not unfunded on paper: adopted, and absent. A rider today is using a network that is well under half finished, and the finished part is nine tenths paint.

None of those 1,976 miles is on this campus, and none of them ever will be. The scheme grades public streets, and the university's interior is not a public street. Below is the same commute drawn as one line, which is the shortest way to see what that means.

The graded network carries a rider to the gate and stops. California grades public streets, and a university's interior is not in the scheme, so the rider crosses one line, leaves the classified network, and only then meets the strictest rule on the whole trip. 20 of our 31 respondents did not know that rule existed. A commute guide can tell you which class of lane carries you to the gate. Nothing on the graded network tells you the last quarter mile is closed.

Stop 03 · The gap

Built for cars, banned for bikes

Every campus writes rules for micromobility. Far fewer build for it. That asymmetry is the whole story. A university can ban a behavior with a sign and a policy number tomorrow. Providing for it takes racks, corrals, charging, and protected room to move, so provision lags restriction by years.

Rules are cheap and provision is expensive, so a campus manages micromobility mostly by limiting it. The result reads as hostility to riders. It is really just the cheaper half of the job.

Benchmark · The peer set

Other campuses already built it

UC San Diego is not short on ideas or money. It is short on provision. Put it beside its peers and the gap is plain: campuses of similar size and climate have spent years turning micromobility into infrastructure, not just policy. The status column below is the campus rating in the League of American Bicyclists' Bicycle Friendly University program.

UC San Diego holds Bronze, the entry rung of the League's four-level ladder. UC Davis and UC Santa Barbara sit at Platinum, the top. The lesson is not that UC San Diego bans too much. Every campus here bans something. The leaders also build.

The stakes · Do the math

Every short trip is a choice

The barriers are not abstract. They decide, trip by trip, whether a person rolls or drives. And the trips at stake are small: half of all US trips are three miles or shorter, yet 72% of those short trips are still made by car. Transportation is the single largest source of US greenhouse gas emissions, at 28% of the total. Short car trips are the most replaceable miles we have.

The University of California knows this. In 2013 it became the first big US university system to promise systemwide carbon neutrality by 2025. In July 2024, a year before that deadline, it dropped the target and reset to a harder one: cut emissions at least 90% from 2019 levels and fully decarbonize operations by 2045, with milestones in 2030, 2035, and 2040. Commute miles are exactly the kind of emission those goals now have to reach. A campus that made riding easy would be doing climate policy, not just managing traffic.

Stop 04 · Field site

One campus, up close

We did not have to look far for a case. UC San Diego is dense, hilly, and fed by the Blue Line trolley extension that opened in 2021, which drops thousands of people a short, awkward distance from where they actually need to be. That last stretch is exactly what a scooter is for, and much of the campus solves it that way.

The university manages the crowding with policy. Riders on shared paths are capped at 8 mph, roughly double walking speed. Weekday riding on Library Walk is prohibited from 8:30 a.m. to 5 p.m., when the corridor fills with pedestrians. Devices stay out of buildings and off the Geisel Library forum level entirely.

Those rules exist because riders and walkers are pushed into the same space. In our observations, the near-misses cluster where a bike or micromobility lane feeds straight into a pedestrian path, like the stretch in front of the art gallery, where riders speed up at the entrance and few have had any formal training. Speed caps and dismount zones treat that friction. They do not design it out.

Stop 05 · How we know

How we ranked it

The ranking is not a guess. It rests on three methods that each cover the others' blind spots: a survey of riders and non-riders, now at 31 responses across twelve institutions and four countries, seven on-site interviews across three campuses, and direct observation of where micromobility thrives and where it stalls, counted on one shared form at two campuses that sit on opposite ends of this argument. The survey is still open, and the figures in the next section are current as of July 25, 2026.

We observe first, because seeing the steep spots and the crowded corridors tells us what to ask. The survey then gives the broad shape across more than one campus and more than one country, since the reasons someone does or does not ride are much the same everywhere. The interviews add back what a survey flattens: a person telling their own story, in their own order.

We ask non-riders as much as riders on purpose. The people who gave up on micromobility, or never started, know its barriers better than anyone still riding through them.

The price · City collision records

Who actually pays when it goes wrong

Everything above is about whether a trip is pleasant, permitted or possible. This section is about what happens to a body when it is none of those things. The City of San Diego publishes every collision its police department records, one row per person involved, with that person's role and how badly they were hurt. Nobody working this corridor has put it next to the policy before, so we did.

We took the four police beats that cover UC San Diego and its approaches, over the eight complete years from 2018 to 2025. That is 3,570 collision reports and 7,145 people. Bicyclists and pedestrians are 452 of those people: 6.3% of everyone involved. They are 36% of everyone who died.

Three in ten drivers are hurt. Nine in ten cyclists are. The gap widens as the injuries get worse: 1.3% of drivers were severely or fatally hurt, against 7.2% of cyclists and 11.3% of pedestrians. A pedestrian in these records is killed at 3.48% against a driver's 0.27%, about thirteen times the rate. This is not a subtle statistical effect. It is the difference between having a steel shell around you and not having one.

2020 is the lowest year in the set and nobody should read that as progress: it is the year the campus emptied. Set it aside and the eight years hold a flat line. Whatever has changed on this corridor since 2018, it has not changed this.

Why this is a climate section, not just a safety one

The case for micromobility is not that riding is fun. It is that a trip taken on a bike or a scooter is a trip that emits nothing, and almost every one of those trips replaces a trip that would have emitted something. That substitution is the entire climate argument, and it is the reason a campus that is 90% cars at its gates is a climate problem before it is a traffic problem.

Our own survey is blunt about how that substitution does not happen. Asked what drives their choice, exactly one respondent in 31 called environmental considerations a major factor. Eleven said they play no part at all. Nobody is riding to save the planet, and a policy built on the assumption that they will is a policy built on nothing.

Which is why the collision records matter here. People choose the mode that feels survivable, and this corridor sends an unmistakable signal about which modes those are. The two modes that emit nothing are the two modes that absorb the harm: 6.3% of the people, 36% of the deaths. A rider who reads that signal and drives instead has not made an environmental choice, they have made a physical one, and the emissions follow. Every unbuilt separated mile and every stretch where the rule says get off and walk is a small tax on the low-emission option, paid in the one currency riders actually respond to. The climate cost of this corridor is not in what it emits. It is in the clean trips it quietly talks people out of.

A further 114 people carry no role in the file or the role “other”. They are counted in every total above and in the extract, and left off the chart because there is nothing to label them. Our extract is published as collisions-corridor.csv and is browsable in the data viewer. It is rebuilt from the city's file by tools/build-collisions.py, and every percentage on this page is recomputed from the counts before each release. The original file is the City of San Diego's police collision details dataset.

The carbon · Our own arithmetic

What one point of mode share is actually worth

The climate case for micromobility is usually made in adjectives. It is worth making in numbers, so here is ours, built from four inputs and nothing else. A typical car emits about 400 grams of CO₂ per mile. The trip is 1.5 miles each way, the leg our own calculator uses. The commute runs five days a week across a 30-week year, which is 150 days. Multiply it out and one person driving that commute burns 450 miles and 180 kg of CO₂ a year.

We do not know how many people drive to this campus, and picking a number to make the total look impressive would be exactly the move this site exists to argue against. So the unit here is a rate: per 1,000 daily car commuters. Those thousand people put 180 tonnes of CO₂ into the air every academic year just getting to class and home again. Scale it to whatever the real figure turns out to be.

The right-hand column is the one that keeps us honest. A shared e-scooter is not zero: manufacturing it, charging it and driving a van around at night to collect it works out at about 220 grams per mile across its whole life, which still cuts the same trip by 45% but not by 100%. A bicycle is the only mode on this page that is genuinely nothing.

Set that against what Bridging the Gap counted at the campus edge, where five of six entry points ran at 89% cars or higher. A ten-point shift is not a utopia. It is one commuter in ten, on a trip most of them could ride in under fifteen minutes, and it is worth eighteen tonnes a year per thousand people who make it. The University of California has committed to cutting emissions at least 90% from 2019 levels and fully decarbonising operations by 2045. Commute miles are precisely the kind of emission that target now has to reach, and they are the kind a campus can move without inventing any new technology at all.

Every figure in this section is recomputed from its four inputs by tools/check-facts.py before each release, and the inputs are in data/facts.json under climate. If a number here disagrees with that arithmetic, the release does not ship.

The same corridor · Three teams

Nobody has measured the thing that matters yet

Three SYN 100 teams have now worked this corridor, none of us together, and each of us stopped at a different point on it. Read side by side they make one continuous route with one gap left in it, and the gap is the part the campus actually controls.

Bridging the Gap took the leg before the gate, and their counts are earlier on this page. We took the leg after it, which is everything above. Between the two sits a third project that went straight at the junctions themselves.

Cyclalytics, by Ian, Frank, Mel and Anyla, is a drone and computer-vision project. They flew a drone thirty to one hundred feet over campus at midday passing period and ran an image-recognition model across the footage to heat-map where riders and walkers collide. They read the same two documents we do, PPM 270-11 and the campus Bicycle, Skateboard and Scooter Guidelines, and reached the same complaint we did: the policy lacks specificity. Their proposal names seven places to point the camera.

An earlier version of this page called their work a proposal rather than a result. That was wrong, and correcting it matters more than tidying it away. They built the instrument and they ran it: four Python tools around a YOLOv11s detector fine-tuned on the VisDrone aerial dataset, ByteTrack to hold each rider's identity across frames, and a smoothing pass that turns thousands of tracked points into a speed heatmap. Their repository ships a labelled drone still and finished heatmaps made from their own footage. That is a working pipeline, not a plan for one.

Every one of the seven is the same kind of place: a surface you are meant to ride on running into a surface you are meant to walk on. That is not a coincidence in their site selection, it is the shape of the campus. Fifteen of our respondents named where the close calls happen for them, and the answers land on those same three kinds of ground: eight said the bike lane, eight said intersections, seven said the pedestrian path.

The valuable thing in their work is not the drone. It is the quantity they named. They call it the slowdown factor: what a rider pays to cross out of the surface built for them and into the one built for walking. They planned to measure it at a single junction by comparing how congested each side was.

What their pipeline does not yet produce is that quantity in units anyone can act on, and they say so themselves. Their speeds come out in pixels/second, not miles per hour. Their README is explicit that converting would require camera calibration and a known ground-plane scale, which is the hard part of aerial measurement and the part a quarter does not leave time for. A pixel is a different distance at every altitude. There is a second limit in their source that is easy to miss: both speed scripts filter to the VisDrone motor class, so the heatmap describes motorbikes and scooters. Their cyclist detection lives in the single-image script, which draws boxes and no speed. So the gap on this corridor is not that nobody built the instrument. Somebody did. It is that the instrument has not been calibrated, and until it is, the slowdown factor has no value in the real world.

Set that term next to the rule this whole site is about, and it does the work for us. A junction imposes a slowdown for a few seconds. The weekday dismount ban does not slow that crossing down. It removes it, for the length of the corridor, from 8:30 to 5, whether anyone is standing there or not. You do not need a drone to measure a slowdown factor of one. That is the number a policy sets by writing it, and it is the reason a rider gets off and walks the last stretch of a trip they made specifically to avoid walking.

Read down the last column and the gap stops being an opinion. Three teams and a municipal dataset cover the trip from a front door to the campus edge and across the roads that surround it. The stretch nobody has an instrument for is the stretch the university itself controls, and it is the same stretch the rule governs.

What the three of us leave behind is a route measured end to end except in one stretch: the last few hundred metres, where the graded network has run out, the rule is strictest, and the only instrument pointed at it reports in pixels.

Stop 06 · Play the barrier

Ride it yourself

The findings are also a game. You ride a scooter down a route modeled on UC San Diego and hit the same three barriers our data points to: nowhere to park, nowhere to charge, and a corridor that closes to riders in the middle of the commute. It is short on purpose, one campus route and a simple score, built by one of us.

You will not have a clean run, and that is the argument. The game is not rigged against the rider. It is built to the specification of a real campus, which turns out to be the same thing.

Stop 07 · The playbook

What a campus can actually do

None of this is exotic, and we rank the fixes by cost and impact so a transportation or housing office could start without a new budget line. In rough order of leverage for the money:

1. Park where people stop, not where it was easy to build. Corrals at dining halls, residence entrances, and transit drop-offs, signed clearly enough that a rider can find one. Mostly paint and racks.
2. Give personal devices somewhere to charge. A few secured, monitored charging points near residence halls do more for adoption than any campaign, and cost far less than the elevator that would otherwise be the fix.
3. Govern speed, do not blanket-ban. Geofenced slow zones in crowded corridors, like the 8 mph shared-path cap, keep the direct route legal instead of pushing riders onto longer detours or into traffic.
4. Treat micromobility as part of transit, not a rival. Racks and charging at every trolley and shuttle stop, so the last mile is not the hard mile.
5. Standardize a little safety. A short required orientation and clear right-of-way where lanes feed pedestrian paths, the exact places collisions happen now.

Ranked by cost and impact, the cheapest moves turn out to be the highest leverage. That is the good news buried in the problem: most of what is missing is provision a campus can already afford.

Stop 08 · Who's asking

Who's asking

This project is not neutral, and we would rather say so. Each of us comes to it from a different side of the same path.

We have all been riders or walkers, never the student who has to drive, or someone for whom an abandoned scooter is not an annoyance but a wall, or the administrator who has to find the money. Those are the views we work hardest not to miss.

Stop 09 · Ride smart

Know the rules before you ride

The fastest way to keep the corridor open to everyone is to ride it the way the campus asks. Most of the friction between riders and walkers comes down to a short list of rules that not enough people have ever been told.

The full rules live in UC San Diego's micromobility policy and the campus police guidelines, both linked in the resource hub below.

Reference · The basics

What counts as micromobility

Micromobility covers small, light vehicles that move at roughly bicycle speed, powered by the rider, a small motor, or both. Owned or shared, docked or dockless, the family looks like this.

California sorts e-bikes into three classes: Class 1 and 2 assist to 20 mph, Class 3 to 28 mph. On campus the 8 mph shared-path cap applies to all of them. The common thread is range: each one covers exactly the under-three-mile trips that make up more than half of American travel, at a fraction of the energy, street space, and cost of a car.

Stop 10 · Take action

What you can do this week

Reading this changes nothing on its own. Here is the shortest path from the page to something on the ground, whether you ride, walk, or just want the corridor to work for everyone in it.

Toolkit · Raise the alarm

Make the case in five minutes

Awareness is not the point. Provision is. Here is everything you need to move this past your own screen: a letter the deciders will actually read, a flyer for a real wall, and a link built to be shared.

Email the deciders

Written and ready to send. Copy it, add your name, and send it through the Transportation Services contact page linked in the resource hub. Ask for the two cheapest fixes: parking where students stop, and somewhere to charge.

To: UC San Diego Transportation Services and Housing Subject: More bike and scooter parking and charging on campus Hello, I am a member of the UC San Diego community who gets around campus by bike or scooter, and I am writing to ask for two low-cost improvements. First, please add secure parking where students actually stop: at dining halls, residence entrances, and the trolley and shuttle stops. Where racks exist they are often full or hard to find, so devices end up in the way. Second, please add a few secured charging points near the residence halls. Many students cannot charge a personal e-scooter or e-bike where they live, and that single gap keeps devices parked and people in cars. Both are cheaper than the workarounds they replace, and both move the campus toward its climate goals. Thank you for considering them. Sincerely, [Your name]

Copy the email

Post the flyer

A print-ready poster with the story, the three barriers, and a code that opens this site. Put it where the bikes actually park.

Open the printable flyer →

Share the ride

The fastest argument is letting someone lose the game. Send the link, or copy this page to drop in a group chat.

Send the game → Copy this page link

Resource finder · Every UC

Find your campus

This project is UC San Diego's, but the pattern is systemwide. Pick any University of California campus to pull its official micromobility resources: where to ride, where to park, how to register a device, and what the local rules are. Every link goes to the campus itself.

The resource hub

Every UCSD micromobility resource, in one place

If you take one thing from this project, take this list. Everything a rider, a walker, or an administrator needs to understand, ride, fix, or push to improve micromobility at UC San Diego, plus the wider context it sits inside.

Reference · Questions

Common questions

Short answers to what riders, walkers, and administrators ask most. Every answer draws on the rules and sources listed on this page.

Reference · Glossary

The words we use

Micromobility

Small, light vehicles that travel at roughly bicycle speed: bikes, e-bikes, e-scooters, skateboards and the like, powered by the rider, a small motor, or both.

Mode share

The percentage of trips taken by a given way of getting around. A campus where 15 in 100 trips are by bike has a 15 percent cycling mode share.

Last mile

The short, awkward final stretch of a journey, from a transit stop to the actual destination. Too far to walk comfortably, too short to drive, and exactly what micromobility is best at.

Provision

What a campus builds and offers: parking, charging, lanes and lighting. The opposite of restriction, and the half of the job this project argues gets skipped.

Geofencing

Using GPS to change how a shared device behaves inside a mapped zone, for example automatically slowing a scooter to a walking pace in a crowded corridor instead of banning it outright.

Dismount zone

An area where riders must step off and walk their device, such as Library Walk during weekday hours.

E-bike classes (California)

Class 1 and Class 2 assist up to 20 mph; Class 3 assists up to 28 mph. On campus the 8 mph shared-path cap applies to all three.

Credits · The team

Who made this

A SYN 100 project for the Synthesis Program at Seventh College, UC San Diego. We came to it as riders and walkers on the same crowded paths, which is the point of view, and the limit, we work from.

• Yash Goel rode Spin and a personal scooter for three years and hit the parking and charging walls firsthand.
• Yuvia Cabral rode shared scooters for two years and brings the safety and convenience lens.
• Rut comes to it as a pedestrian and argues for accountability, not just access.