Home > Portfolio >

Laboratory Robot - automated process design

First concept design for discussion.

Links:

Project description | Concept drawings | Components & process | Choice of equipment | Control Logic | Input/Output definitions

Project description

To reliably automate part of the sample preparation and analysis process and free a person to do other things.

• A technician loads a magazine within a freezer with syringes containing the relevant sample.
• A robot arm takes the syringe to a station where chemicals are added by an external machine.
• The arm transfers the syringe to a sonic device for pulverising.
• The arm moves the syringe above a magazine loaded with lidded sample bottles.
• A bottle is lifted onto the syringe needle.
• A nozzle is lowered into the top of the syringe and gas used to force the syringe contents into the sample bottle.
• The syringe and needle are stored for recycling.
• A second robot arm presents the sample bottle to the chromatograph.
• After sampling the bottle is deposited in a freezer and the entire process starts again.

The process needs to be "fail safe" or "no fail" so each stage of the process is confirmed by a proximity sensor. Should any confirmation fail, the process will halt and signal a fault.

Concept drawings

As a way of discovering what the project actually invloves I have drawn a couple of portions of the installation - ARM A and the magazine inside FREEZER 1. This has enabled me to extrapolate and describe the rest of the installation.

The detail shown should be sufficient to permit discussion and if the remainder of the installation is drawn in similar detail, dimensions can be added and working drawings submitted for accurate quotations.

Definitions of major components and
a description of the process

ARM A contains a number of input and output devices to accomplish a series of tasks:
• A gripper lined with rubber to hold the syringe throughout the process. The gripper is spring-loaded open and closed by a servo. [Image]
• A mechanical proximity switch is mounted below the gripper to confirm the presence of the syringe. If at any stage this switch opens, the entire process halts and the controller signals a fault.
• A carriage holds gripper, servo and switch. The carriage can extend and retract (nominally 100mm) driven by a linear actuator and position defined by proximity switches. [Image]
• The head holds the carriage and can be raised and lowered (approx. 100mm) by an actuator, positioned by switches. [Image]
• The whole assembly is supported by the truck which travels along a track, driven by an actuator, positioned by switches. [Image]

POSITION 1 - FREEZER 1 contains a motor driven endless chain with 40 magazine slots. [Image] Each slot is preloaded with a syringe, spigot and needle.
On activation, the motor runs for 3(?) secs or until the body of the syringe assembly opens a proximity switch.
The switch confirms the presence and correct pickup position of the syringe.
If the prox does not confirm within the time, the process halts and the controller signals a fault.
ARM A extends and grasps a syringe, retracts, and tracks to Position 2.

POSITION 2 - FLUID INJECTION uses an external robot device to add chemicals to syringe which is held in a programmed position.
ARM A extends the syringe within range of a sensor (unknown type - research required) which confirms sufficient liquid is in the syringe.
ARM A retracts and tracks to Position 3.

POSITION 3 - PULVERISING uses a sonic probe mounted in a fixed position.
ARM A extends and raises so the probe is immersed. Probe switches on.
ARM A lowers to a mid position and raises again. Repeat as required. Probe off.
ARM A down to low position, retracts and tracks to Position 4.

POSITION 4 - TRANSFERRING SYRINGE CONTENTS to sample bottle. There are several components in Position 4.

• the Magazine - sample bottles are pre-loaded into a magazine similar to that in Position 1. The ready bottle is lifted onto the syringe needle by a piston from below.
• the Spigot Rotator - considerable force is needed to turn the tap on the syringe assembly so ARM A will push the spigot tap into a molded wheel and simultaneously the spigot inlet and outlet tubes into molded sockets. The in/out tubes get additional support from a gripper which closes over the back of the spigot. When the wheel turns, the tap turns with it - hopefully. (They are very stiff!)
• the Nitrogen Injector - the nozzle is mounted in a raised position, high enough to allow ARM A to present the syringe assembly below it, and directly above the ready sample bottle. The nozzle is raised and lowered by an actuator, positioned by switches.
• ARM B is a servo driven gripper mounted at the end of a rigid arm which pivots to several known positions. Motor driven pivot with proximity switches.

The Magazine rotates the next bottle to the ready position.
ARM A extends. The syringe fits into the Spigot Rotator.
An additional gripper closes around the spigot inlet and outlet.
A piston lifts the bottle between the jaws of the ARM B gripper, pushing the needle through the rubber top.
The ARM B gripper closes firmly. A prox between the gripper jaws reports the continuing presence of the bottle.
The Nitrogen Injector lowers the nozzle into the top of the syringe.
The Spigot Rotator turns 90°.
Nitrogen on and off.
The Injector raises to up position.
The additional gripper retracts from the spigot neck.
ARM A raises, retracts and tracks to Position 5.
The Magazine piston retracts.
ARM B rotates and presents the sample bottle to the external robot serving the chromatograph.
A sensor (type unknown - research required) reports liquid in bottle.
After sampling ARM B rotates to a third position - Freezer 2 - where the gripper opens and the sample bottle is dropped into a hatch into Freezer 2.
ARM B rotates to the home position.

POSITION 5 - SYRINGE RECYCLING is an open hopper which directs the used syringe assemblies into a container.
ARM A gripper opens and releases the syringe assembly.
ARM A tracks to Position 1 and resets.
On a signal from the chromatograph the process repeats.

Choice of equipment:

In general I am proposing fairly old but very reliable technology - worm drives and mechanical switches - in preference to software controlled stepper motors, even though the overall process is software controlled.

There are several reasons:
• a worm drive will hold its position even without power
• a switch provides a positive check of a mechanism's position where a stepper motor counts steps to establish position and could conceivably hit an obstruction which throws the count off;
• this is a one-off design to accomplish one process - the flexibility of software controlled steppers is not required;
• I have considerable experience with older technology and less with steppers.

The equipment suggested is either of industrial standard or is highly reliable and in common use. Without knowing your budget I have stayed away from wiper motors which while very reliable and high torque need a fair amount of workshop time to mount and machine - the mounts are usually at odd angles and positions and a special socket has to be made to accommodate the wiper spindle. And there is no warranty if they are used in a non-specified operation. I have also stayed away from hobby motor/gearboxes - the motors supplied are subject to poling and need replacing immediately; the bearings and teeth are sloppy making them less than reliably accurate, and the units are noisy.

Servos. Servo motors as used in radio-controlled models are comparatively cheap, readily available, robust, have high torque, can be set precisely and are feedback controlled to ensure a constant correction for any change in position. Eg: If the syringe assembly is bumped, the servo will squeeze the gripper more tightly to maintain its grasp.

Mechanical proximity switches. Double pole mini switches are durable, cheap, replaceable and easily obtained.

Motors & gearboxes. These, or similar, small DC drives and gearboxes can be purchased with specific specifications and are easily mounted.

Linear Actuators. A DC motor spins a collar which winds up and down a threaded shaft. The actuators are precise and robust with a capacity up to 160kg. By mounting the moving component in a track system, movement can be confined to precise straight lines rather than traveling through an arc.

Control. The initial design of the inputs and outputs indicates that we will require at least 25 out channels and more than 29 inputs. It could be done with a PIC or PLC although the largest PLC with which I am familiar is 32 channels. We could link 2 of these together - not ideal - or source a larger module. A PC is still required to make any changes and every change has to be uploaded - a time consuming process which can be very frustrating during initial programming and testing.

My preferred option is to install a digital I/O card in one of the slots in a PC and run the whole process from there. I have used these controllers previously in installations which have been running now for 5 plus years without a glitch. They are driven by 486-100s and while one has a new mother board (it got "moused" - rodent pee is corrosive!) and both have had the hard drives replaced, the cards are fine.

There are any number of cards available. These are from Advantech:
• to suit ISA slot - "...a full-size PC add-on card that offers 144 programmable digital I/O lines divided into six main ports. Each port in turn consists of three subsidiary 8-bit ports. You can individually configure each of these ports as input or output..."
• to suit PCI slot - (this project would require 1 card only) "...a 96-bit digital I/O card for the PCI bus, which can be extended to 192 digital I/O channels by connecting with its extension board. The 96 I/O lines are divided into twelve 8-bit I/O ports. Users can configure each port as input or output via software..."
• either card will require 2 relay card for switching motors and other outputs - "...provides 16 SPST power relay channels with a max contact rating of 250VAC @ 5A or 30VDC @ 5A and can be driven directly by the digital output from PC-LabCards..."
Virtually any card uses proprietary software. The "Genie" program is somewhat clunky to use but appears to be completely stable. A first draft of the control logic indicates no obvious difficulties.