Think in parts deliver in goosebumps

The Purdue Undergraduate Rocket Propulsion Lab (PURPL) exists to put real propulsion hardware in student hands, and TURBOPUMP is its flagship: a 5,000 lbf open-cycle kerosene-LOX turbopump and rocket engine, single-shaft and turbine-driven, designed to spin at roughly 50,000 RPM. Fluid dynamics, thermodynamics, rotordynamics, mechanical design and manufacturing, converging in one compact machine drawn by undergraduates.

The CAD was ready. Turning it into flight-grade metal was the problem PURPL brought to Imaginarium, and it opened our first official aerospace case file.

Five components, two alloys: the gas generator injector, kerosene casing, LOX casing and LOX impeller grown in SS316L, the kerosene impeller in AlSi10Mg. The impellers carry complex turbomachinery blading; the casings hide intricate internal passages and interfaces that have to survive machining and assembly. Nothing here only had to match an outer envelope. Every as-printed surface had to protect the bolt patterns, mating faces and machining operations coming after it.

Our involvement began before production. We reviewed the designs with PURPL’s Turbopump team while they were still moving: build feasibility and orientation, internal supports, accessibility of complex geometry, material selection, machining allowances, critical interfaces and tolerances. Manufacturing constraints entered the design directly instead of surfacing later as failures, cutting off the design, manufacture, discover, redesign loop before it could start.

The parts were printed near-net shape on purpose. PURPL’s machining plan called for roughly 0.04 in of allowance on critical surfaces, with bolt patterns held to defined positional requirements, so the as-printed geometry banked material exactly where the team’s own mills would need it. We printed; PURPL machined in-house, keeping control of every critical interface while we handled the geometry only metal AM can produce.

Collaboration ran through production, Mumbai to West Lafayette. Dimensional inspection and build feedback moved between the teams, so the as-built condition was judged against requirements together, including working through internal supports and printability changes on the LOX casing before it hit the build plate.

The full component set is delivered and moving through post-machining and integration at Purdue. The next milestone is testing: this hardware is headed for a hot-fire campaign. PURPL is an exceptionally competitive, technically ambitious group of engineers, and we are proud to be in their corner when they light it.

aerospace CAD becomes
flight-ready hardware
within 2 weeks.

Ascento AG, an ETH Zurich spin-off, builds autonomous robots that patrol outdoor industrial sites around the clock, through rain, heat, dust and the occasional knock. The head is where the hardest problems live: it carries the sensing and vision stack, takes the brunt of the environment, and defines the robot’s face. Building it demanded 4 different manufacturing disciplines. Most suppliers do one well. Ascento needed all 4, without fragmenting their supply chain across 4 vendors.

We manufactured the structural core as one integrated build: CNC machined components, precision sheet metal and welded sub-assemblies. Tight tolerances where the sensor stack mounts, robust fabrication where the environment hits hardest, and a finish that stands up to Swiss engineering scrutiny.

Protective bumpers vacuum cast in polyurethane across multiple shore hardnesses, tuning impact absorption zone by zone. The bumpers were overmoulded on a sheet metal frame as well to provide more strength in certain areas.

We also vacuum casted the front orange light cases with pigmented transparent polyurethane to ensure functionality of the end use parts

For the face shield guarding the vision systems, we moved to production tooling: injection moulded polycarbonate with the optical clarity the camera stack demands and the impact and weather resistance a 24/7 outdoor patrol robot requires.

Our team worked as an extension of Ascento’s, Mumbai to Zurich: DFM feedback on geometries, material selection across metals, polyurethanes and thermoplastics, tolerance and fit resolution across processes, and honest advice on when to prototype, when to cast, and when to invest in tooling.

we built one robot head, not four purchase orders.

An injury took a fragment of a patient’s skull, and the bone was too deformed to reuse. The conventional fix, a hand-ground titanium mesh, meant opening the skin just to measure the damage, then days of grinding, fitting and cost the patient could not afford.

We sat with the surgeons and rebuilt the missing skull in CAD from X-ray and CT data, then grew the implant in Ti-6Al-4V titanium on DMLS. The same 3D model let the doctors rehearse the surgery before a single incision.

anatomy became the CAD.

Tethys Robotics AG, a Zurich-based ETH Zurich spin-off, builds autonomous underwater drones for missions too dangerous, too deep or too murky for human divers: search and rescue in fast-flowing rivers, infrastructure inspection in zero-visibility water, operations in contaminated environments. Underwater is one of the most hostile places a machine can be sent. Constant immersion, corrosion, pressure, impacts against debris, and no tolerance for failure mid-mission.

Tethys needed a complete vehicle body at pre-production volumes without steel tooling, corrosion protection engineered into the parts themselves, stiffness without weight, and propeller guards that flex without fracturing. And they needed all of it from one partner who could reason about materials, geometry and manufacturability as one problem, not four.

We produced the full drone body in vacuum cast polyurethane, with pigmentation integrated directly into the casting, so colour is part of the material, not a surface layer that peels or fades underwater. Specialised coatings protect the body against corrosion and degradation through sustained immersion, a hull that holds its integrity and appearance over a long operational life. Production-quality bodies at exactly the volumes a scaling robotics company needs, with no tooling investment locked in before the design matured.

Strength underwater cannot come at the cost of weight. We manufactured the carbon fibre tubes that run through the drone’s architecture, the stiffness and structural backbone the vehicle needs, while keeping mass, and therefore buoyancy and power demands, tightly controlled. Carbon fibre’s immunity to corrosion makes it a natural match for permanent underwater service.

The propellers are the drone’s most exposed and most essential components. We engineered their protective covers, the “ears” of the vehicle, in high shore hardness polyurethane: rigid enough to shield the propellers from strikes and debris, flexible enough to absorb impact energy and spring back rather than crack. Two properties that normally fight each other, resolved in one material. A guard that protects without shattering, mission after mission.

Across every workstream, our engineers worked alongside the Tethys team: DFM feedback on castability and part geometry, polyurethane system and shore hardness selection, pigment and coating specification for marine endurance, and integration guidance so cast, composite and machined elements come together as one sealed, serviceable vehicle. One partner reasoning about the whole machine, not a vendor quoting part numbers.

built for a world that destroys hardware.

Bionic hands cost a fortune and arrive in standard sizes. With Makers Hive, we set out to build KalArm, India’s first fully functional 3D printed bionic arm, at a price patients could actually reach.

MJF-printed PA 12 body: biocompatible, lightweight, durable and custom-fit to every wearer, then painted and finished in post so it looks as human as it grips. 40% cheaper to develop, 30% faster to ship.

we printed a hand that fits its human.

A UAE energy client had a water-cooling pump fail on its instrument air compressors. The OEM had moved on: no spares, no CAD, not even an assembly blueprint. The traditional route to remake it: 7 to 8 months and upwards of USD 60,000, for one obsolete part.

We scanned the worn sample into a precise digital twin, then played material detective: fingerprinting the original alloy and matching it to Stainless Steel 17-4 PH for the body and Inconel 718 for the impeller.

Form, fit and function proven in white resin prototypes first, iterated on desktop printers in hours, so every correction happened before a single gram of metal was melted.

Complex geometry grown on metal printers, intricate cuts and sealing faces machined after printing, simpler components straight to CNC. Geometry from the printer, precision from the mill.

the spare part became a file.

“The product and turnaround time of 3 weeks for the parts is quite impressive.”