Featured

Refrigerated Automated Storage & Dispensing Robot

Cartesian style robot, with dynamic storage arrangements and patented product manipulation

Sector: Automation
Years: 2023–2025
Engagement: In-house
Role: Lead Engineer & Primary Mechanical Design Engineer
Status: Prototype validated; four pre-production units built
Client: Future Robotics

Patent PCT/EP2025/055523; A Robotic Food Vending System

SolidWorks · Sheet metal design · Manufacturing drawing · Siemens PLC (basic troubleshooting and test) · Machine building and testing · R&D · Supplier Liaison · Management & Coordination

Problem

Conventional refrigerated vending fixes slot size, so unit capacity is set by the largest product it must hold. The commercial driver here was capacity, specifically for meal sized boxes across roughly three size classes. The brief was a machine where any slot accepts any product within a defined envelope, packing arrangements dynamically, handled by a single gripper head rather than size specific mechanisms.

Constraints

External envelope fixed by the top mounted refrigeration housing; not alterable as the design evolved
Internal shelf and cell layouts fixed early in the project
Product envelope 170 × 90 × 400 mm (W×H×D)
Single two sided gripper head handling every size class with one operation
Pre-production redesign had to retain prototype mounting points, cutouts and interface geometry
Full load from empty in under 20 minutes; the primary driver of ballscrew lead and motor selection
Customer dispense of a complete 2 to 3 item order in under 60 seconds
Products pre-sealed, so hygiene scope was routine internal cleaning; all shelves stainless steel and removable for washdown

Approach

Owned all mechanical architecture: mechanisms, ergonomics, safety, maintenance access, cable routing, material selection
Reversed the packaging strategy for V1.1; auxiliary components placed in the model before carriage chassis geometry was fixed, where the prototype had let structure drive component placement
Designed as a hierarchy of independent sub-assemblies, allowing parallel design, build and rework
Modelled all electrical and pneumatic components inside the mechanical assemblies to verify clearance and routing before build
Sequenced sub-assembly design by estimated supplier lead time, releasing long lead items first and replanning continuously around delays, test results and revisions

Gallery

Section view rendering showing double sided product manipulation head and custom mini conveyor inside unit(Render)

Image of same robotic head sub assembly (cover removed) outside of unit(Photo)

Rendering of robotic head sub assembly as part of greater 3 axis robotic assembly(Render)

Custom compact conveyor design, used for product lateral positioning(Render)

First iteration of the gripper head sub-assembly — Four iterations of same sub assembly across prototypes to end result

Second iteration of the gripper head sub-assembly — Four iterations of same sub assembly across prototypes to end result

Head assembly, prototype iteration — Iteration of entire head assembly between prototype and pre-production machine

Head assembly, pre-production iteration — Iteration of entire head assembly between prototype and pre-production machine

Animation from early design phase of prototype to demonstrate unique product loading concept

Y axis, 4 timing belt drive gearbox and assembly (removed)(Photo)

Head lower sub assembly with air lines plumbed(Photo)

Outcome

The prototype validated the manipulation concept. Four pre-production units followed, with improved robustness and adjustability, faster assembly, and pneumatic routing concealed within the head envelope. Product height capacity rose from 75 mm to 90 mm and ballscrew extension time fell from roughly 30 seconds to 5. The prototype took four months to build; the four pre-production units took three, assembled single-handed. Development ended when the company ceased trading.

Testimonial

“He has a very inventive mind and is always able to find a solution to any problem that comes his way.”
See full testimonial
Gerry Horan
Director & Owner (Retired) · Horan Automation Ltd & Future Robotics Ltd

Further detail

D1Z-axis actuation: prototype to pre-production

Three targets came out of prototype testing: speed, product height (75→90 mm), and stroke. V1 used a laterally mounted stepper driving through a belt. V1.1 moved to a custom higher lead ballscrew with direct drive, removing belt compliance, with an upgraded motor sized for the inertial load. Result: extension time fell from roughly 30 seconds to 5, closing the load cycle time constraint this was solving.

D2Component first packaging strategy

In V1, structure drove component placement. In V1.1, components were placed first and the chassis built around them. That reversal produced defined clearances for moving and serviceable parts, symmetrical load paths, better adjustability, and improved assembly access.

D3Compact conveyor for confined space

A standard conveyor was infeasible in the envelope discovered during prototype testing. Footprint outside the belt face was minimised, the servo relocated beneath the belt where headroom existed, a compact timing belt drive fitted, and adjustability retained for alignment.

D4Building four machines alongside the design

Limited external machine build support meant assembling four complete machines while still designing. The modular sub-assembly hierarchy allowed staggered build across supplier lead times. Fastener selection was driven by tool access and serviceability, not load alone. Slots and adjuster plates absorbed tolerance variation between outsourced and fabricated parts. An improvement log was kept from day one of assembly, and every entry was implemented in V1.1.