Repair or replace? How AI decides what your car needs
Every Chassly damage report makes a repair-vs-replace call for each affected part. This decision often surprises owners: a small-looking dent can recommend full panel replacement, while a much bigger dent might be flagged for repair. The logic isn't arbitrary; it follows a few specific rules grounded in material science, modern manufacturing, and shop economics. Understanding the rules helps you read your report critically and have informed conversations with body shops.
The material rule: steel forgives, aluminum doesn't
The single biggest factor in repair-vs-replace is what the panel is made of. Steel panels, found on most economy cars, trucks, and SUVs, can be reshaped by skilled body technicians using paintless dent repair (PDR), conventional dent pulling, or filler and refinishing. Steel is forgiving: it can be massaged back into shape without compromising structural integrity.
Aluminum panels, found on Tesla, many BMW models, Audi, Range Rover, F-150 (since 2015), and most luxury European cars, are much less forgiving. Aluminum work-hardens when struck, meaning the metal becomes stiffer and more brittle in the deformed area. Trying to reshape it often leads to cracks, paint failure, or weakened structure. Beyond minor dents (under 2 inches with no paint damage), aluminum panels are almost always replaced.
Chassly's AI knows which manufacturers use aluminum extensively and adjusts its recommendations accordingly. A 4-inch dent on a 2018 Honda Accord (steel) might recommend repair; the same dent on a 2018 Tesla Model 3 (aluminum) almost certainly recommends replacement.
The severity threshold: 6 inches is the rough cutoff
Independent of material, dent size correlates with repairability. Most steel panel dents under 6 inches across, without paint cracking, can be repaired with PDR or conventional bodywork. Above 6 inches, the panel typically has structural deformation rather than just surface dent, and replacement becomes more cost-effective than the hours of labor required to reshape and refinish.
Paint damage shifts this threshold. Once paint has cracked, the entire affected area needs sanding, priming, base-coating, clear-coating, and blending, which is hours of skilled labor. A 4-inch dent with cracked paint sometimes costs more to repair than a 5-inch dent with intact paint, because the paint work is the expensive part.
Glass and lighting follow different rules. Cracked glass (windshields, side windows, headlight lenses) is almost never repaired; they're replaced as full assemblies. The exception is small windshield rock-chips under the size of a quarter, which can be resin-injected. Anything larger requires a new windshield.
Functional vs cosmetic: safety always replaces
If damage affects function (a headlight that won't aim correctly, a bumper that won't hold its mounting points, a hood that doesn't latch, a door that doesn't seal), the AI always recommends replacement. Trying to repair structural or functional damage creates safety risks that aren't worth the cost savings.
This is also where insurance adjusters and Chassly tend to agree: insurers are far less willing to cover repaired parts that affect crash safety. If a fender is deformed enough to potentially redirect crash energy in a future collision, the actuarial answer is replacement.
Examples of functional damage that always triggers replacement: cracked headlight or taillight lenses, bent door frames that prevent sealing, hood damage near the latch or hinges, bumper covers that no longer secure to crash bars, broken side mirrors with internal mechanisms exposed.
The economic rule: when repair costs more than parts
There's a hidden economic threshold body shops use: when estimated repair labor exceeds about 60% of replacement part cost, replacement wins. The math: 8 hours of body work at $90/hour is $720 in labor. A new fender at $250 plus 2 hours installation at $90 is $430 total. Even though the new fender costs less than the repair labor, you're getting a factory-perfect part instead of skilled hand-bodywork.
Chassly's cost estimator factors this in. When you see 'repair' on a part with a higher cost range than replacement would imply, it's because the part has limited replacement availability (custom paint, special trim, ordered-from-Japan part) and repair is the only practical option.
On older vehicles, the calculation flips. A 2008 Civic fender might be repairable for $400 in body work; a replacement part with paint is also $400 but has a 2-week lead time. Repair wins on speed even when costs are comparable.
When you should question Chassly's recommendation
AI recommendations aren't gospel. Get a real body shop estimate when: (1) the recommendation is replace but the part looks salvageable to you, (2) the cost is high enough that a second opinion is worth the time (anything over $1,000), (3) you have a body shop relationship and want their input, (4) you're dealing with insurance and need professional documentation regardless.
When you bring photos to a body shop, sharing your Chassly report can speed up the conversation. The shop already has the AI's analysis as a starting point and can either confirm or push back on specific calls. Most shops appreciate the engaged customer: fewer surprises during the repair.
Pro tip: shops are sometimes more aggressive about replacement than repair because replacement is more predictable. They know exactly how many hours a part swap takes; PDR can take 30 minutes or 4 hours depending on how cooperative the metal is. If a shop quotes replace and the AI suggests repair, ask 'is repair an option?'. Often it is, just less convenient for the shop.
Frequently asked questions
Why does the AI sometimes recommend replace but my friend's shop says repair is fine?
Skilled body shops can sometimes repair damage that the AI's conservative threshold flags for replacement. Conservative recommendations exist because repair quality varies dramatically by shop, and the AI optimizes for outcomes most owners will get from typical shops. A great shop in your area might exceed those typical results.
Can I always insist on repair to save money?
On steel panels with cosmetic damage, often yes. On aluminum or carbon-fiber panels, no: the limitation is physical, not preference. On safety-critical or functional damage, also no. Even if a shop is willing, you're creating a problem for your next collision.
How accurate is the AI's material judgment?
Very accurate for major manufacturers and current models. The AI knows Tesla uses aluminum extensively, Toyota largely uses steel, Audi uses aluminum on premium trims, etc. For low-volume or specialty vehicles, the AI may not know and tends to recommend conservatively (replace) when uncertain.
Does insurance typically agree with repair-vs-replace recommendations?
Insurance adjusters use industry estimating tools (Mitchell, CCC, Audatex) that often align with AI recommendations. Disagreements happen when the adjuster's quick visual inspection differs from the per-panel analysis the AI performed; sharing your Chassly report can support your case if you want a specific outcome.