Paul Johnson v. Ridge Tool Manufacturing Company, Inc.

In the
United States Court of Appeals
For the Seventh Circuit
____________________
No. 25-2604
PAUL JOHNSON,
Plaintiff-Appellant,
v.
R
IDGE TOOL MANUFACTURING COMPANY, INC.,
Defendant-Appellee.
____________________
Appeal from the United States District Court for the
Northern District of Illinois, Eastern Division.
No. 1:21-cv-01939 — LaShonda A. Hunt, Judge.
____________________
ARGUED FEBRUARY 24, 2026 — DECIDED AUGUST 7, 2026
____________________
Before BRENNAN, Chief Judge, and EASTERBROOK and
TAIBLESON, Circuit Judges.
BRENNAN, Chief Judge. Paul Johnson severely injured him-
self using a drain-cleaning machine manufactured by Ridge
Tool Manufacturing Company, Inc. He filed suit, but the dis-
trict court excluded his expert witnesses and granted sum-
mary judgment to Ridge Tool. He appeals.

2 No. 25-2604
We affirm. In line with Rule 702 and our gatekeeping role
for opinion evidence, Johnson’s experts were appropriately
excluded for using unreliable methods and reaching unsup-
ported conclusions. And without that evidence, his suit can-
not succeed. Thus, the district court properly granted sum-
mary judgment.
I. Background
Ridge Tool Manufacturing Company, Inc. designed and
manufactured the K-6200 drain cleaning machine. On a
wheelbase sits a large drum, which holds about 100 feet of
coiled 5/8-inch cable. When turned on, the drum feeds out ca-
ble at 285 revolutions per minute. The operator guides the
spinning cable into the drain. To reduce the potential for user
injuries, Ridge Tool took several precautionary measures.
For one, sometimes the machine cannot be placed near the
drain’s inlet. The greater distance between the machine and
the inlet, the more cable is exposed. That can “result in [the]
cable twisting, kinking or control problems,” the user manual
warns. So as a solution, Ridge Tool offers an optional front
guide-hose that channels the exposed cable into the drain,
sparing the user from handling the moving cable.
For another, Ridge Tool includes specific gloves, offered in
one size, to use with the machine. The manual instructs users
to wear them while operating the machine. Otherwise, a
“loose fitting cloth glove,” the manual cautions, could “be-
come wrapped around the cable and cause serious injury.”
And on the machine is a warning label. At the top, it de-
picts a mangled hand with a cable coiled around it, with the
statement: “Cables can twist, kink or break. Rotating parts can
strike, catch, cut and crush body parts.” Below that, the label

No. 25-2604 3
instructs users to read the manual to understand the potential
risks of injury.
In August 2019 Paul Johnson used the K-6200 to clean a
drain. He neither read the manual nor wore the gloves Ridge
Tool provided, opting instead for his employer’s pair, despite
them being “a little large.” He did, however, read the warning
label on the machine.
Johnson started the machine and began guiding the cable
into the drain. As cable fed out, he supported it with his hand.
But then, the middle finger of his glove “just got wrapped up
into the rotating cable and it just took me along with it.” John-
son’s arm and body were twisted to the right, severely injur-
ing him.
Johnson sued Ridge Tool for strict liability, negligence,
and failure to warn. On the first two theories, he alleged the
company manufactured an unreasonably dangerous and de-
fective machine. As Johnson saw it, Ridge Tool should have
installed a fixed hose-guard, which would have made holding
the bare cable in his gloved hand unnecessary. For support,
he relied on an expert, Chad Jones, a mechanical engineer.
Jones’s report proposed that an alternative fixed-guard de-
sign that, in his opinion, would have made the machine safer.
Under a third theory, Johnson alleged Ridge Tool failed to
warn users of the known hazards associated with the ma-
chine. For support, he relied on another expert, Cynthia
Rando, a human factors professional.
1
Rando’s report largely

1
A human factors expert examines how people interact with technol-
ogy, products, services, and environments to ensure optimal safety.

4 No. 25-2604
opined on the deficiencies with the user manual and the risks
of the glove’s one-size design.
Ridge Tool asked the district court to exclude the reports,
opinions, and testimony of both experts and moved for sum-
mary judgment. The court granted both motions.
The court first excluded both expert witnesses under Fed-
eral Rule of Evidence 702. In its view, Jones’s evaluation was
unreliable. His opinion—that there was a safer alternative de-
sign that Ridge Tool should have used—was unsupported be-
cause, as the court recognized, he failed to test his alternative
design to ensure it was indeed safer. Nor did Jones address
aspects of the alternative design: its compatibility with the
machine, the added manufacturing costs, or its risks. “With so
many unanswered questions here,” the court reasoned “it is
guesswork as to whether a fixed guard would make a drain
cleaning machine safer.” The second expert, Rando, was ex-
cluded for different reasons. She focused largely on the user
manual and its perceived shortcomings, which Johnson did
not read. Nor did she offer examples of what a more effective,
alternative warning label would be.
The district court also granted Ridge Tool’s summary
judgment motion. Johnson’s negligence and product liability
legal theories principally relied on the now-excluded expert
reports. Without evidentiary support, the theories failed. Nor
could Johnson’s failure-to-warn theory succeed. Because the
Rando report was also excluded, “the record [did] not contain
any admissible evidence that the warning was inadequate or
an alternative warning would have been better.” In any event,
even if the Rando report had been admissible, Johnson would
have found it unhelpful: Rando “never opined on these par-
ticular topics in connection with the warning label, and, as

No. 25-2604 5
[Ridge Tool] points out, the warning label instructed users to
read the user manual, which [Johnson] admittedly failed to
do.” Johnson appeals.
II. Federal Rule of Evidence 702
A. Governing Legal Principles
On the admissibility of experts, courts walk a tightrope.
On one hand, judges are gatekeepers to ensure junk science
does not reach the jury. Experts in necromancy or astrology
have no business in the courtroom. See Kumho Tire Co. v. Car-
michael, 526 U.S. 137, 151 (1999). Our role is to “ensure that
any and all scientific testimony ... is not only relevant, but re-
liable.” Daubert v. Merrell Dow Pharms. Inc., 509 U.S. 579, 589
(1993).
On the other hand, courts should not subsume the jury’s
“essential role as the arbiter of the weight and credibility of
expert testimony.” Stollings v. Ryobi Techs., Inc., 725 F.3d 753,
765 (7th Cir. 2013). The crucible of cross-examination should
reveal the expert’s blemishes. “Determinations on admissibil-
ity should not supplant the adversarial process; ‘shaky’ expert
testimony may be admissible, assailable by its opponents
through cross-examination.” Gayton v. McCoy, 593 F.3d 610,
616 (7th Cir. 2010).
Federal Rule of Evidence 702 strikes this balance. As
amended in 2023, it states that a witness may testify as an ex-
pert “if the proponent demonstrates to the court that it is more
likely than not that: (a) the expert’s scientific, technical, or
other specialized knowledge will help the trier of fact to un-
derstand the evidence or to determine a fact in issue; (b) the
testimony is based on sufficient facts or data; (c) the testimony
is the product of reliable principles and methods; and (d) the

6 No. 25-2604
expert’s opinion reflects a reliable application of the princi-
ples and methods to the facts of the case.” FED. R. EVID. 702.
In short, a witness is an admissible expert if they are qualified,
speak to relevant issues, and employ reliable methodologies.
Gopalratnam v. Hewlett-Packard Co., 877 F.3d 771, 779 (7th Cir.
2017).
And courts must be mindful of Rule 702’s 2023 amend-
ments. First, “expert testimony may not be admitted unless
the proponent demonstrates to the court that it is more likely
than not that the proffered testimony meets the admissibility
requirements set forth in the rule.” F
ED. R. EVID. 702 advisory
committee’s notes to 2023 amendment. Second, “expert opin-
ion must stay within the bounds of what can be concluded
from a reliable application of the expert’s basis and method-
ology.” Id. The court’s duty to shield the jury from “expertise
that is fausse and science that is junky,” is now unmistakable.
2

Kumho Tire, 526 U.S. at 158–59 (Scalia, J., concurring).
Appellate courts give district courts broad discretion in
Rule 702 decisions. We review de novo whether the district
court applied Rule 702, asking whether the court “provided
more than just conclusory statements of admissibility or inad-
missibility to show that it adequately performed its gatekeep-
ing function.” Anderson v. Raymond Corp., 61 F.4th 505, 508
(7th Cir. 2023) (citation modified). If, for example, the court’s

2
On December 1, 2023, Rule 702’s amendments went into effect. The
district court here issued its summary judgment ruling in August 2025.
Because the amendment’s rules were in effect at the time, we consider
them here. Cf. In re Onglyza (Saxagliptin) & Kombiglyze (Saxagliptin & Met-
formin) Prods. Liab. Litig., 93 F.4th 339, 345 n.4 (6th Cir. 2024) (“We employ
the old rule because it was still in force at the time of the district court’s
decision.”).

No. 25-2604 7
consideration is limited to a single sentence, no deference is
due. Id. at 508–09. But if the district court stated the requisite
elements of Rule 702 and offered reasoned consideration—
not just conclusions—we review for abuse of discretion. Timm
v. Goodyear Dunlop Tires N. Am., Ltd., 932 F.3d 986, 994 (7th Cir.
2019). Here, the district court’s analysis spanned several
pages, correctly identified Rule 702’s elements, cited cases, ex-
amined the record, and addressed counterarguments. That is
more than sufficient for us to review for abuse of discretion.
B. Jones’s Report
In Illinois, a plaintiff’s case against a product manufac-
turer is supported by showing that a safer alternative product
design was feasible. Winters v. Fru-Con Inc., 498 F.3d 734, 744
(7th Cir. 2007). Here, Johnson offered Jones’s opinion that a
safer alternative design to the K-6200 was “technologically
and economically feasible”: a coil spring guard fixed to the
machine that guides the cable into the drain, making it unnec-
essary to touch the moving cable.
Yet Jones conducted no tests nor ran any models to prove
the design’s feasibility. And in alternative design cases, we
“have recognized the importance of testing.” Cummins v. Lyle
Indus., 93 F.3d 362, 368 (7th Cir. 1996). But to Johnson, this is
not a problem because Jones pointed to competitor machines
that already have the alternative design—those machines, he
says, are testing the design for him.
True, we have recognized that an expert need not test an
alternative design that is already in the marketplace. Consider
Anderson v. Raymond, 61 F.4th at 505. There, a worker fell out
of a forklift, which then ran over her leg. Id. at 507. The plain-
tiff’s expert opined that the forklift should have had a door,

8 No. 25-2604
despite not testing that alternative design. Id. at 510. That
omission was not necessarily fatal to the admissibility of his
opinion, Anderson explained, because the company itself sold
a forklift with a door: “[W]hen data are available from another
source, there’s no need to duplicate that information by test-
ing.” Id. As we said, “Raymond’s customers who have elected
to fit their forklifts with Raymond’s optional door have been
testing [the expert’s] alternative for him.” Id.
Hillman v. Toro is another example. 164 F.4th 971 (7th Cir.
2026). There, we again rejected the need for testing an alterna-
tive design because it was already in the marketplace. Id. at
982. The plaintiff’s expert “relied on zero-radius-turn mowers
offered by other companies that included independent
brakes ,” meaning the “alternative designs were taken from
comparable products actually on the market.” Id.
Anderson’s principle, Johnson argues, means Jones did not
need to test the alternative (fixed coil guard) design. To John-
son, the fact that three competitors’ machines include such a
guard is a sufficient basis to admit his expert report.
We disagree. To be sure, testing an alternative design is
unnecessary if it is already used in the marketplace, as Ander-
son and Hillman instruct. But those cases do not hold that just
pointing to competitor products is enough to admit an ex-
pert’s opinion. That a design exists, without more, cannot be
the sole basis for an expert’s opinion. Ye t that describes
Jones’s report. To him, three comparators prove that an alter-
native design is feasible, so Ridge Tool should have incorpo-
rated that design when manufacturing the machine.
The inadequacy of Jones’s report is evident when com-
pared to the expert reports in Anderson and Hillman. In

No. 25-2604 9
Anderson, the court noted the expert had, “reviewed troves of
data generated from forklift accidents,” analyzed “tip-over”
injury data, and studied design changes made by the forklift
manufacturer. 61 F.4th at 510. In Hillman, the expert pointed
to four or five competitor products using the safer alternative
design. 164 F.4th at 981. But the expert also cited several in-
dustry publications and patents to support and explain the
need for an independent brake. Id. at 981–82. And he quoted
a report given to the American Society of Agricultural Engi-
neers discussing the need for independent braking. Id. at 982.
Jones’s report reads much differently. He failed to apply
the factors collected in Cummins, 93 F.3d at 362, as the district
court pointed out. For example, Jones did not evaluate “the
short- and long-term maintenance costs associated with the
alternative design” or “the degree to which the alternative de-
sign is compatible with existing systems and circuits.” Id. at
369. Nor, the district court noted, did he consider that touch-
ing the cable may improve the machine’s safety because users
receive tactile feedback. In addition, Jones cites no relevant
publications to conclude that a fixed guard is important or
necessary. And his report fails to address the relative safety
of the proposed alternative design and the actual machine.
Surely this machine’s injury rate must be known; it could be
one (J ohnson) or much higher. But without that figure, the
machine’s safety relative to others is inconclusive. An expert
opinion requires some basis, but Jones had none.
Nor is Jones’s comparison to the competitor products all
that probative. One way to identify an unreasonably safe de-
sign is whether it “conform[s] with the design standards of
the industry.” Anderson v. Hyster Co., 74 Ill. 2d 364, 385 N.E.2d
690, 692 (Ill. 1979). Jones pointed to three other competitors

10 No. 25-2604
that used a fixed guard. The district court asked the right
question: three out of how many machines? Three out of four
with a fixed guard likely means that design is the industry
standard; but three out of four hundred proves little. A driver
who hurts his head on the roof of his Camry cannot point to a
Porsche as proof that convertibles are the industry standard.
Jones lacks the denominator, which the district court correctly
held against him.
But what about the optional front guide-hose that Ridge
Tool offers with the machine, Johnson points out. Because
Ridge Tool already has a guide-hose design, a fixed design is
thus feasible. But the designs are different. That the machine
has one type of guard is not proof that another type is feasible.
Thus, Ridge Tool’s own design is not evidence that a different
design is an industry standard.
In sum, Jones’s methodology amounted to a faulty com-
parison and little else to reach his conclusion. So, t he district
court did not abuse its discretion when it excluded his report.
C. Rando’s Report
Johnson’s second expert, Cynthia Rando, is a certified hu-
man factors professional of about twenty years. The district
court excluded Rando’s report and testimony for two reasons.
First, Rando opined that the manual was unclear, as it pro-
vided conflicting instructions. Yet Johnson had not read the
manual.
3
So, he cannot establish that the alleged flaws caused
his injury. See Solis v. BASF Corp., 2012 IL App (1st) 110875,
979 N.E.2d 419, 439 (Ill. App. Ct. 2012) (“[A] plaintiff who
does not read an allegedly inadequate warning cannot

3
ECF No. 63-1 at 63.

No. 25-2604 11
maintain a product liability action premised on a failure-to-
warn theory.” (citing Kane v. R.D. Werner Co., 275 Ill. App. 3d
1035, 657 N.E.2d 37, 39 (Ill. App. Ct. 1995)).
Recognizing this problem, Johnson argues that the warn-
ings on the machine were insufficient, and he cherry-picks
statements from Rando’s report to support this theory. But as
the district court accurately found, Rando did not offer an
opinion about the warning label. Even reading her report
charitably, not much can be extracted about the machine’s
warnings. Her report does not even include a picture of the
warning label.
The district court recognized a second deficiency with
Rando’s report: it does not propose an alternative warning
that she believes would have prevented Johnson’s injury.
Without discussing what designs would have effectively
warned a user, any opinion about the ineffectiveness of the
current warning is “akin to ‘talking off the cuff’ and not ac-
ceptable methodology.” Bourelle v. Crown Equip. Corp., 220
F.3d 532, 539 (7th Cir. 2000). Though no single factor is dis-
positive under Rule 702, an expert’s failure to “design[] or
suggest[] to the court an alternative warning that would have
been appropriate” signals unreliability. Dhillon v. Crown Con-
trols Corp., 269 F.3d 865, 870 (7th Cir. 2001).
Bourelle is illustrative. There, plaintiff brought a failure-to-
warn theory after being injured by a forklift. 220 F.3d at 533,
538. Her expert failed to test or even to draft an alternative
warning. Id. at 538. That rendered his opinion “unreliable” for
not using an “acceptable methodology.” Id. at 539. So too here.
The court excluded Rando’s report as she failed to “draft or
test any proposed alternative warnings.” Id. Rando, in other

12 No. 25-2604
words, is merely speaking offhand, so excluding her report
was not an abuse of discretion.
III. State-L aw Violations
Now to the merits. As mentioned, Johnson sued for negli-
gence, product liability, and failure to warn. The district court
granted summary judgment to Ridge Tool on all three theo-
ries. We review that decision de novo, and all inferences are
drawn in favor of Johnson. Giovannelli v. Walmart Inc., 164
F.4th 1052, 1054 (7th Cir. 2026).
We can quickly dispense with the negligence and product
liability theories.
4
On appeal Johnson suggests these theories
rise or fall with the admission of Jones’s report. He did not
argue those theories are viable regardless of whether the re-
port is admitted. By contrast, he did claim that his failure-to-
warn theory is viable with or without Rando’s report. Nor
does he independently analyze them beyond a request to re-
verse the grant of summary judgment.
With the failure-to-warn legal theory, there is no waiver
problem. In this diversity case, Illinois law applies. Giovan-
nelli, 164 F.4th at 1054.
Illinois holds manufacturers liable for failing to warn of a
product’s dangers. The product must have a “latent defect or
a dangerous propensity of which a user would normally be
unaware,” and the manufacturer has failed to provide an ad-
equate warning of the danger. Hakim v. Safariland, LLC, 79

4
These are legal theories, not claims. “A claim is the set of operative
facts that produce an assertable right in court and create an entitlement to
a remedy. A theory of relief is the vehicle for pursuing the claim.” St. Au-
gustine Sch. v. Underly, 78 F.4th 349, 352 (7th Cir. 2023).

No. 25-2604 13
F.4th 861, 868 (7th Cir. 2023) (quoting Baltus v. Weaver Div. of
Kidde & Co., 199 Ill. App. 3d 821, 557 N.E.2d 580, 588 (Ill. App.
Ct. 1990)). An inadequate warning, for example, may fail to
caution about the specific “risk presented by the product” or
“reach foreseeable users.” Id. at 869 (quoting Palmer v. Avco
Distrib. Corp., 82 Ill. 2d 211, 412 N.E.2d 959, 964 (Ill. 1980)).
Johnson has little evidence to support his failure-to-warn
theory. Rando’s opinion was properly excluded. But even if
not, much of Rando’s report was irrelevant to the warning la-
bel. Although Johnson’s legal theory can likely proceed with-
out an expert, id. at 871, he is left with (1) the warning itself
and (2) his testimony describing how the machine injured
him. That thin evidence is insufficient to survive summary
judgment.
Moreover, Johnson offered no alternative warnings—la-
bels that, had they been affixed to the machine, would have
adequately alerted him to the machine’s dangers. Illinois
courts have ruled that in failure-to-warn suits, alternative
warnings are necessary to survive summary judgment. “To
prevail,” a plaintiff “must establish that greater warnings
would have prevented [the] injuries.” Muhammad v. Abbott
Lab’ys, Inc., 2022 IL App (1st) 210478, 203 N.E.3d 1001, 1018
(Ill. App. Ct. 2022). Said another way, plaintiffs cannot defeat
something with nothing. “In failure-to-warn cases ... there
must be sufficient evidence supporting a reasonable infer-
ence, rather than a guess, that the presence of adequate warn-
ings would have prevented the plaintiff’s injuries.” Solis, 979
N.E.2d at 439 (citation modified).
Johnson reads our decision in Hakim as greenlighting fail-
ure-to-warn cases that have no alternative warning evidence.
In Hakim, this court considered whether expert testimony was

14 No. 25-2604
necessary to prevail on failure-to-warn theories in Illinois. 79
F.4th at 869. The answer depended on the specifics of the
product. In cases involving drugs or medical devices, “expert
testimony may be necessary for a jury to determine how a rea-
sonable specialist in the relevant field would interpret the
warning.” Id. at 870. But other times, the product’s dangers
are apparent. Id. at 871 (door-breaching shotgun rounds do
not disintegrate when they hit wood). For the latter, expert
testimony for the failure-to-warn theory is unnecessary. Id.
From Hakim’s holding, Johnson extrapolates that if expert
testimony is not always necessary, and experts introduce
alternative warnings, then alternative warnings must not al-
ways be necessary. To Johnson, his failure to offer an alterna-
tive warning, then, is not necessarily fatal.
We doubt that Hakim can be read that broadly. Cf. Elion v.
United States, 155 F.4th 889, 905 (7th Cir. 2025) (rejecting an
expansive reading of precedent); Olivier v. City of Brandon, 607
U.S. 552, 566 (2026) (overbroad sentence in prior case “was not
meant to address” certain suits). Hakim stands for the rule that
in Illinois, an expert witness is not always necessary in failure
to warn cases. Inferring more from that case incorrectly ex-
pands its holding, particularly because Illinois law seems to
require alternative warnings. See Muhammad, 203 N.E.3d at
1018; Dhillon, 269 F.3d at 869, 871 (excluding expert for not
offering alternative designs and then granting summary judg-
ment because plaintiff “does not dispute” that “without the
proffered expert testimony, he could not prevail”).

No. 25-2604 15
But even if not strictly necessary, without an alternative
warning it can be difficult to establish causation.
5
See Walker
v. Macy’s Merch. Grp., Inc., 288 F. Supp. 3d 840, 867 (N.D. Ill.
2017) (Without alternative warnings, there is no evidence that
the plaintiff’s “injuries would ... have happened ‘but for’ the
lack of warnings on the [product].”). Take this case as an ex-
ample. Without an alternative warning, we are left guessing
whether a better one would have prevented Johnson’s acci-
dent. It is not obvious he would have understood the purport-
edly clearer warning. After all, a more descriptive, and thus
longer, label may have discouraged Johnson from reading it.
In the end, all Johnson offered was the warning label on
the machine and his assertion that it was deficient. But a party
needs more than conjecture and conclusions to defeat sum-
mary judgment. In re Abbott Lab’ys, Preterm Infant Nutrition
Prods. Liab. Litig., No. 25-2587, 2026 WL 2137765, at *5 (7th Cir.
2026). Johnson’s lack of evidence cannot overcome summary
judgment.
* * *
The district court did not abuse its discretion when exclud-
ing Johnson’s two experts, and it correctly granted summary
judgment.
A
FFIRMED

5
See Solis, 979 N.E.2d at 439 (collecting case to show that both strict
liability and negligent failure-to-warn legal theories require proximate
causation).