Cambridge is the hand-over point — the Vision Park drop-off, or the post. From there, every drive walks the same benches by the same route, whether it arrived from Cambridge, Newmarket or Huntingdon. Consider this the tour: bench by bench, what each piece of kit is, what it does, and why it settles which recoveries work.
Intake write-blocking, imaging done in hardware, firmware repair, and a clean-air corner for the mechanical jobs — each job runs through the stations it needs and skips those it doesn't.
The single most useful thing to understand about how this work differs from what you can do at home.
When your computer reads a sector and fails, it retries. It retries hard, repeatedly, because its job is to return that data and it assumes the disk is basically healthy. On a failing drive that assumption is wrong, and those retries are what finish it off — hours of the head repeatedly seeking to a damaged area it cannot read.
A hardware imager does the opposite. It reads once, notes the failure, moves on, and comes back to the difficult areas afterwards with different timing and different settings. It can be told to skip ahead, to read in reverse, to change how long it waits before giving up.
That difference is why an imager frequently recovers 99% of a drive that ordinary copying kills at 12%. Same disk, same damage, opposite approach.
Purpose-built acquisition hardware that reads damaged media without hammering it, with control over timing, retry behaviour and head selection.
A filtered environment for head replacement and platter work. Drives run with sub-micron head clearances; a single dust particle under a head scores the surface.
Manufacturer engineering interfaces for repairing the service area modules a drive loads at start-up, and for reaching SSD controllers that have stopped responding.
Programmers and test-point rigs for reading NAND directly, then reversing the wear levelling, interleaving and scrambling the controller was applying.
A hard drive is a small computer. It boots from code held in a reserved area of its own platters — the service area — loading modules that describe its geometry, its defect map and the calibration values for its specific heads. If those modules are damaged, the drive spins perfectly and reports nonsense, and no amount of software running on your computer can touch it.
Adaptive data is the other half of that. Each drive is calibrated at manufacture to its own heads and platters, and those values live in a ROM chip on the circuit board. Fit a board from another drive and the heads are being told to behave in ways that do not match the hardware, which is why board swapping fails and why the chip has to be transferred instead.
Both are routine work here and neither is reachable by anything you can install.
Only if it has to be opened, which is mechanical-class work. Electronic, firmware and logical faults never require it, and those are the majority.
On a healthy drive with a logical fault, frequently yes, and we will tell you so. On a drive that is physically failing, software keeps it powered and seeking for hours, which is the load that finishes it.
Because a hardware imager reads once and moves on rather than retrying until the drive gives out, and because it can reach firmware areas your computer never talks to.
No. Everything is done at Vision Park by our own engineers, including clean-air work and chip-level reads. Data does not leave the UK.
Held for 14 days after delivery in case anything was missed, then securely wiped unless you have asked for it back or asked us to destroy it sooner.
No. A drive fitted with donor parts is a vehicle for one read. It is not a disk to trust with anything afterwards.