MithralonEvery machine will have a record.
Mithralon is building a cross-manufacturer operating record for robot fleets. Repairs and software changes go into one history that insurers and lessors can review when a machine changes after it was insured or leased. The sample record is fictional. The questions behind it are not.
One software change open for review. v4.4.0, planner policy update. The manufacturer's regression suite passed; the test of the machine's movement limits at the site is still pending.
Written for your file, not for the machine's dashboard.
When a robot changes after it was insured or leased, through a software update or a new task, how does the party carrying the risk find out, and what does finding out cost? That is the question we are putting to insurers and lessors right now. The sample record shows one possible answer. If any line of it would enter your file, that is worth twenty minutes.
The EU Machinery Regulation 2023/1230 applies from 20 January 2027.
One history across manufacturers.
A manufacturer's records cover its own machines. An insurer pricing a mixed fleet works from a loss history that is still short. And a lessor taking back a three-year-old robot sees whatever paperwork the operator kept. The information is thin exactly where capital wants to flow.
Mithralon is meant to close that gap from the ground up. Brand, fault, cause, fix, part, downtime. One entry per resolved incident, across manufacturers.
The Fleet Evidence Console (demo) is a separate fictional example of how a software update opens an evidence review.
Fault code in. Guided fix out. One more entry in the history.
A record has to come from somewhere. Our working assumption is that it starts on the shop floor, with the technician. A fault code goes in, a plain-language cause and guided steps come out, across brands, and one tap confirms what worked. That confirmation is the entry. No telemetry and no IT project are needed for it.
Entries are kept apart by what they can show. A repair report says what a technician reported. A manufacturer's release note says that a version exists. A version report from the machine says which version a unit is actually running. A review says what someone checked and what is still open. The record is meant to label each entry as one of these, and to leave a gap visible where one kind of evidence is missing instead of filling it with a claim.
Fleets are mixed. Knowledge is locked inside each brand.
A modern site runs transport robots from one manufacturer, arms from a second, cleaning robots from a third. Each brand ships its own portal, its own error language, its own hotline, its own parts logistics. The knowledge is organized vertically, brand by brand. The reality on the floor is horizontal, site by site.
When a robot stops with a fault code, a clock starts. The technician digs through PDFs, waits on hotlines, calls the colleague who saw this code once before. The machine stands still the whole time. The core problem is not that robots fail. Failure is physics. The problem is the time lost between fault and fix, and that time is a knowledge problem.
Machines, priced against their record.
Cars were financed and insured for decades before their histories became knowable. Our thesis is that robot fleets get their history layer much earlier, because leasing models and high unit costs pull the financial questions to the front.
A record written with technicians is meant to become the standard that maintenance, parts, insurance, leasing and resale price against.
Personally built. Personally answered for.
I'm Max Stuffler. Mithralon is pre-product. I'm building Mithralon from Saxony, in Europe's largest robot market, under the EU data law that gives fleet operators access to their machines' data. I write the code myself and I take every early conversation personally.
Twenty minutes, nothing to buy.
Insurers, lessors, brokers
A conversation about the last real case where technical evidence about a machine mattered to a submission, a quote, a claim or a residual value.
Talk to the founderOperators and service teams
A conversation about how you handle downtime today. Fleets of roughly twenty to two hundred robots across two or more manufacturers, and the workshops that keep them running. In return you see the technician workflow first and shape what it covers.
Talk about your fleet