Introduction
A resistograph does not show you the inside of a tree. It records resistance along a single line — the path of a 3 mm needle — and nothing more. That limitation is not a flaw. It is the reason the tool works: the drill takes a hypothesis you built by looking at the tree and pushes it close to proof.
The instrument I use is the IML-RESI PowerDrill, built by the German manufacturer IML. The question I hear most often from arborists overseas is whether a resistograph finds cavities. It finds the cavities that sit on the line you drilled. Everything below is an explanation of that sentence, written from the perspective of someone who runs an IML-RESI on street and park trees in Tokyo, inside a diagnostic system that is more prescriptive than most.
In Japan, the Drill Only Appears at Stage Three
The four stages Tokyo puts in writing
The Tokyo Metropolitan Government’s Bureau of Construction publishes a Street Tree Diagnosis Manual (gairoju shindan-to manyuaru, 街路樹診断等マニュアル), and a diagnosis moves through four stages:
Tree inspection → Visual diagnosis → Instrument diagnosis → Overall judgment
What makes the manual unusual is that it also assigns each stage to someone. Tree inspection runs on a routine cycle and does not require a certified specialist — it is the screening layer. From visual diagnosis onward, the work belongs to a jumoku-i, and the interval to the next diagnosis is set by the result of the last one — anywhere from one to ten years. Instrument diagnosis happens only on the trees the visual stage identifies as needing it. The final judgment is made by a person weighing everything together, not by a graph.
The 2021 revision also changed the vocabulary: what had been called “initial diagnosis” became “tree inspection,” and “precise diagnosis” became “instrument diagnosis.” The emphasis moved with it — from extracting unhealthy trees to inspecting in order to keep trees healthy.
The judgment is a schedule, not a verdict
The overall judgment sorts each tree into one of four classes: A (sound), B1 (minor abnormality), B2 (moderate abnormality), and C (serious abnormality). What matters is what each class commits you to. A tree judged B2 receives a follow-up diagnosis around a year later — a re-examination written into the system rather than left to anyone’s memory. A tree judged C moves promptly to replacement or to physical measures. Follow-up diagnosis for B2 trees was added in the 2021 revision, and as far as I know there is no equivalent obligation in the standards used in Europe or North America.
This changes what a resistograph result is for. The graph does not end an argument. It places a tree on a schedule. A number that would be a verdict in another framework becomes, here, a date for the next visit — and that is a far healthier relationship to have with an instrument.
So the instrument comes out of the case late in the sequence, and that is the point. Skip the visual stage and you have no basis for deciding where to drill. Put a machine that sees a single line into a trunk without that basis, hit sound wood, and you have manufactured false reassurance — the most dangerous output a diagnosis can produce.
What actually decides the drilling point
My drilling points come from the abnormalities the visual stage found: an anomalous sound under the mallet, a fruiting body, a wound or a change in the bark. Where those signs appear, that is where the needle goes.
The purpose is specific — to locate the places most likely to lead to whole-tree failure or branch failure. It is not to scan the tree. More holes do not mean more accuracy; they mean more wounds and a longer report. The quality of the visual diagnosis becomes the accuracy of the instrument diagnosis, which is why the field methods that require no technology at all remain the foundation of everything that follows.
Why the operator matters as much as the instrument
In Japan, diagnosis is carried out by a jumoku-i (樹木医) — a nationally certified tree doctor. The instrument sits underneath that qualification, supporting the observer rather than replacing them. The thinking is close to Visual Tree Assessment (VTA) as practised in Europe and North America, but Japan also codifies who performs the diagnosis and in what order — a difference I have written about in more detail in this comparison of the two systems.
The Two Values an IML-RESI PowerDrill Records
Drilling resistance (green) and feed force (blue)
The IML-RESI PowerDrill records two values simultaneously as the needle advances — and recording both, rather than resistance alone, is the specification that matters most in daily use:
- Drilling resistance (green on the graph) — the resistance met by the rotating needle
- Feed force (blue on the graph) — the force required to push the needle forward
Most explanations of how a resistograph works stop at the first. In the field, the second is often the one that tells you something.
Why the feed curve finds decay earlier
IML’s own technical documentation explains the mechanism: shaft friction has a large influence on drilling resistance, but only a minimal influence on feed force.
That difference decides what you can see. The denser the wood, the more shaft friction accumulates along the needle, and the more easily an early decay signal is buried inside the green curve. The blue curve does not carry that burden. When the needle meets a defect, feed force drops clearly, and a stretch where feed force approaches zero indicates a cavity or advanced deterioration.
For hardwoods — which is most of what I measure — reading the two curves against each other, rather than reading the green one alone, is the difference between catching decay early and confirming it late.
Measured, not estimated
This is where the IML-RESI differs fundamentally from the other instrument in the case. Sonic tomography infers internal condition from sound velocity between sensors and interpolates the space in between to produce an image. You get the whole cross section, but what you are looking at is a calculated estimate — as I found when combining tomography with rope access during an aerial diagnosis at a Tokyo temple.
The IML-RESI records actual measured values along the needle path. The view is narrow — one line — but on that line, nothing is inferred. When trees I have measured are later felled and the cross section exposed, the agreement between the drill profile and the real cavity is very high.
Narrow, but certain. That is the character of this machine.
Settings, and How to Read the Graph
Feed speed and needle speed: a starting point, not an answer
Settings change with the hardness of the material. These are the values we use as a starting point:
| Material | Feed speed | Needle speed |
|---|---|---|
| Hard wood | 25–50 cm/min | 2,500 rpm |
| Soft wood | 50–100 cm/min | 2,500 rpm |
I want to be blunt about that table: those are starting values, not fixed settings. We change them according to conditions on site. Moisture content, the stage of decay, the season, and the part of the tree being measured all shift what works. If you take those numbers home and use them unchanged on every tree, you will not get this machine’s accuracy.
Push the feed speed too high and the gradual decline of early decay is compressed on the graph until it is unreadable. Too low and you gain nothing but heat, needle wear, and lost time. Choosing a speed by reading the material is itself an extension of visual diagnosis.
Three patterns worth recognising
- Cavity — both curves fall to the floor and stay flat
- Early decay — a stepped decline from the sound-wood baseline; the blue curve moves first, and more clearly
- Sound wood — a maintained level, with periodic waves corresponding to growth rings in ring-porous species such as keyaki (Zelkova serrata)
Where the graph will mislead you
A completely dead tree. This one deserves particular care. Wood in a dead tree dries and becomes simply hard, and the graph can return high resistance values that resemble sound wood. High numbers do not mean sound wood. The interpretive framework was built for living trees; do not apply it unchanged to a dead one.
The first few centimetres. Bark and cambium make the entry unstable. Do not read that noise as a defect.
A single line. One line is one line. Drill from more than one direction, and check the result against the visual findings.
Sound wood is not the same as a safe tree. Adequate residual wall thickness means nothing if the root system or the growing position is the real problem — a distinction that matters enormously for the street trees of Tokyo, where the trunk is rarely the weakest part of the system.
Visual prediction and drill results agree — but not for everyone
The prediction you build from an anomalous mallet sound, a fruiting body, or a change in the bark agrees with the IML-RESI result a high proportion of the time. The more trees you inspect, the higher that rate climbs.
What I do not want lost in that statement is this: it is not something anyone can do. Distinguishing an abnormal sound requires having heard a great many normal ones. Recognising a bark change as significant requires knowing what unremarkable bark looks like on that species, at that age, in that position. Without accumulated experience, no hypothesis forms — and without a hypothesis, nothing tells you where to put the needle.
The accuracy of instrument diagnosis is set by the quality of the observation that precedes it. The resistograph is not the thing that is good. When the visual diagnosis is good, the resistograph looks good.
What Japan’s Conditions Change
A humid climate, and an annual load test
Japanese summers are hot and humid, and decay fungi stay active over a long season. Then the typhoons arrive and apply a load test to every weakened tree in the city, every year. Where decay advances quickly and the structure is tested annually, the demands on both the frequency and the precision of diagnosis differ from those in a dry region.
Different species, different curves
Keyaki, ichō (Ginkgo biloba), kusunoki (Cinnamomum camphora), Somei-Yoshino cherry — Japan’s principal street and park trees do not match the dominant species of Europe or North America in hardness, or in how decay enters them. Ring-porous and diffuse-porous woods produce different waveforms, and an identical drop in the curve can carry a different meaning.
So the settings, and the drilling decisions, rest on species-specific accumulation. The machine is the same everywhere; the reading is local. What an arborist elsewhere can take from this article is not my numbers, but the structure: you need to build the same body of experience for the species you actually work on.
Frequently Asked Questions
Q: Can a resistograph tell me how big a cavity is?
A: It tells you the length of the cavity along the line you drilled. It tells you nothing about how far that cavity extends across the rest of the cross section. For that, sonic tomography is the appropriate tool. They are not competitors; they answer different questions.
Q: Does drilling harm the tree?
A: The needle is 3 mm. A healthy tree compartmentalises a wound of that size. But the load accumulates with every hole, which is why narrowing the drilling points through visual diagnosis ultimately protects the tree.
Q: Should I read drilling resistance or feed force?
A: Both, in relation to each other. In dense wood especially, feed force — which is barely affected by shaft friction — reveals early defects sooner.
Q: Can I use a resistograph on dead trees?
A: Yes, but the reading changes. Dead wood dries and hardens, so a high resistance value does not indicate sound wood. Do not apply living-tree baselines to a dead tree.
Q: Resistograph or sonic tomography — which should I buy first?
A: They do not replace each other, and a proper diagnosis combines them. But if I had to choose one, I would buy the resistograph first — in my case an IML-RESI PowerDrill. It gives you actual measured values rather than an estimate, and its agreement with the real internal condition is very high. If the budget is genuinely tight, though, invest in visual diagnosis skill before either machine.
Conclusion
A resistograph is not a tool for passing judgment on a tree. It is a tool for moving a hypothesis closer to certainty. The machine does not choose where to drill; a person does, and that decision rests on listening to the mallet, finding the fruiting body, and reading the form of the tree.
Nor should a diagnosis end with the resistograph alone. The drill sees a line. Sonic tomography sees a plane, by estimation. Visual diagnosis sees the tree. Combining them is the shortest route to a correct diagnosis — and I would argue it is also the right one. The moment you hand your judgment to a single instrument’s numbers, what you are doing stops being diagnosis.
Japan’s urban tree management developed inside a particular set of conditions: a humid climate, annual typhoons, its own species mix, and a public manual that specifies the order of diagnosis. The IML-RESI is the same machine everywhere in the world. How it is used is decided by the ground beneath it.
Note: this is not a sponsored article. I write here as a working user of the IML-RESI PowerDrill, and the observations above come from my own measurements on Japanese trees.
Field notes from Japan’s trees arrive a few times each month — subscribe via the form in the footer, or follow the work on LinkedIn.
