Automating Your Lab
Understanding why you need automation for your lab and how you will implement and maintain the system is key to a successful automation project. We’ve put together the most important questions you should be asking yourself as you plan and prepare for building an automated system in your lab.
Getting Started With a Lab Automation Project
As you explore bringing automated systems into your lab, first make sure you've considered these three aspects of your project:
- Establish project goals. Define your specific desired outcomes and benefits from your planned automation project.
- Your application and objectives should dictate robotics and automation. Though you may be more familiar with molecular biology techniques, you may be less aware of robots’ capabilities. If you can manually carry out a workflow with pipettes, filter-columns, centrifuges, heaters/coolers and shakers, then you can also automate the workflow with conventional liquid handlers. If you are unable to automate with conventional equipment, then your project objectives can guide your decisions about project costs versus return on investment.
- Balance cost with return on investment. Any workflow can be automated but developing custom robotics or automation systems will be costly and time-consuming. If the ROI cannot be justified, be ready to adjust your requirements and expectations as you scope the costs and limitations of your project.
To help you navigate these factors, we've put together eight key questions to ask yourself as you begin your lab automation project.
Questions for Starting an Automation Project
Why do I need automation?
Before you start this project, define where you want to go. Are you trying to grow your core facility or business? Do you want your lab to be equipped to handle increased demand? Do you need to automate processing for many different sample types and protocols or just one?
If you do not know where you are going, any road will take you there.
Clearly defining what you need from your automated system will not only make the project run more smoothly, but it will also ensure you get what you need.
An organization wanted an automated DNA extraction system for downstream next-generation sequencing. They needed two instruments for operational redundancy, so they sought to save costs through maximizing efficiency and space utilization on the platform. They decided to purchase two liquid handlers with the smallest capacity that would achieve their sample processing goal. No position was left empty and the throughput per run could not be expanded.
After the organization purchased the instruments and a Field Support Scientist built the scripts and ensured the workflow ran smoothly, the organization realized that their original goal was not specific enough: they needed to extract DNA and RNA from each sample.
To make that switch, the organization had to make serious compromises. Switching to new chemistry and eluting DNA and RNA into separate wells required a different chemistry and substantially more deck space. Consequently, their throughput was reduced by more than half and they needed manual interventions. Their original investment did not meet their needs because they didn't correctly identify their goals when purchasing the new instruments.
After defining a goal, you need to identify the problems you want to solve. Automation can support higher throughput, reduce error, improve consistency, shorten processing times, save costs and more. However, using automation to streamline all your processes could be beyond your budget, so identify and rank your priorities to support better decision-making.
Wondering what can be achieved with lab automation? Most routine workflows have previously been automated. Learn what's possible with support from our automation scientists.
For example, if your lab handles or will handle many different sample types and protocols, then the better choice for you could be purchasing multiple lower throughput systems for specific workflows. If not, then a larger system could be the better choice. Try not to underestimate your needs—purchasing a larger platform is usually less costly than purchasing additional instruments beyond your required redundancy.
Understanding why you need to automate will also help you get your highest return on investment. For example, you might need 6–10 staff to run a semi-automated method on low throughput-bench instruments, whereas a staff of 1–2 individuals can run thousands of samples per day with a higher-throughput setup.
Why Bring Automated Nucleic Acid Extraction into Your Lab?
What processes do you need to automate?
With a defined goal and a list of priorities, you should identify the specific protocols and the associated steps you need to automate. Researching if your manual workflow has been automated can be helpful: ask kit manufacturers or robotics vendors for application notes and more information.
The first step is to map out each step of your desired workflow. Also note which steps of the workflow will be most impacted by automation and what steps will solve specific problems. Your needs and downstream applications will strongly influence the specific workflow. Understanding your sample types, processes and downstream needs is key and will help you better communicate your specific requirements and get the full benefit of your automation investment.
The steps involved with automated extraction of RNA from blood samples will depend on the desired application. For example, consider how an automation protocol changes based on whether the eluted nucleic acids from blood samples are intended for mRNA sequencing or RT-qPCR measurements. Extracting RNA from blood for mRNA sequencing benefits from high sample volumes (approximately 2.5ml) as well as red blood cell lysis and depletion because globin mRNA can consume expensive reads. However, high sample volumes would likely necessitate a costly on-deck centrifuge or manual interventions. Conversely, downstream RT-qPCR experiments may require less material than mRNA sequencing applications, which would allow you to reduce the volumes processed in your automated extraction workflow and to potentially process sample without a red blood cell lysis step. Lower sample volumes can, in general, increase throughput, decrease costs and simplify the workflow.
Consider how an automated nucleic acid protocol may change for a biobank versus a consumer genomics or diagnostics company. For a biobank, maximizing the yield of nucleic acid is important whereas processing times and costs might be less so. A consumer genomics or diagnostics company, conversely, will likely want to drive costs down while only processing enough nucleic acid for typically two-analyses (an initial test and backup). The workflows will look very different.
By mapping out your automated workflow, you can identify the specific steps and equipment needed to achieve that workflow. Here are some specific processes you may need to automate.
- Sample ID management, such as barcoding and tracking.
- Sample preparation.
- Analyte quantification with instruments like spectrophotometers, fluorometer or luminometers.
- Sample normalization.
- Assay setup, including library preparation or drug screening.
You can often find reported automation protocols that use sample types or assays that are relevant to your own project and will help you identify the processes you need to automate. By doing this background research, you can learn what to expect for throughput, required steps, processing time and more. While any single step in these protocols will require a minimum amount of time on an automated platform, they often scale well through parallel processes on larger platforms or by running multiple instruments simultaneously.
As you consider the specifics of your proposed workflow, keep in mind budget, whether you have robotics you can retrofit and if any steps can accept manual interventions.
You should also define the excess capacity you need from your automated system. This will help you plan what size and type of automated system you need.
Your project requirements can and (and sometimes ought to) change as you gain new information, so make sure to prioritize your criteria. In one example, an organization requested a solution for high throughput nucleic acid extraction from 10–40ml of liquid per sample and wanted their workflow to be completely automated. However, automated processing of such large sample volumes would have increased the cost of the consumables and the automation system, driving the cost per sample beyond what the organization could afford. Handling such large volumes would have required costly custom development and the throughput per instrument would have been low, requiring additional instruments to meet their throughput expectations.
However, by reconsidering their method and requirements the project could become financially feasible and still meet their needs. For example, the first step, sample processing, could be carried out manually or pseudo-manually to reduce automation costs. Alternatively, the organization could concentrate samples to reduce volume prior to nucleic acid extraction, though they would have to confirm they could still sensitively detect their analytes. Finally, they could also determine if they can achieve their desired downstream applications with lower volumes.
As you design the specifics of your workflow, it is important to consult with automation experts. Without being fully aware of what automation technologies are available, it is easy to feel constrained by your budget and requirements. However, most basic automation workflows can be achieved with standard robotics systems without custom parts. If you can clearly communicate your exact goals, technical experts can help you identify how to achieve those goals.
Liquid handling platforms from Tecan, Hamilton, Beckman and others are complex to setup and program but will offer the greatest flexibility. These systems excel at moving volumes of less than 1ml and are especially good at rapidly dispensing low volumes for applications like PCR setup and library preparation. They can be adapted to handle larger volumes, but this will often slow processing speeds and reduce throughput.
For nucleic acid extraction, magnetic particle movers are often easier to operate but are also less flexible. For example, ThermoFisher’s Kingfisher systems are very easy to use and automate, but these instruments can only move magnetic particles and mix samples. Other processes, like sample management, reagent dispensing, and assay preparation must be carried out manually or with another integrated automated platform. Your needs will determine if this is sufficient for your lab. Alternatively, Promega’s Maxwell® RSC 48 Instrument uses magnetic particle mover technology with a ready-to-go cartridge-based format to process 1–48 samples per run. Learn more about the Maxwell® RSC 48.
What are your quality requirements?
As you develop your automated workflow, you will need to test its performance and troubleshoot. The following table lists variables that can impact or define sample quality for several application types. Define what your quality needs are so that you can communicate them to an instrument vendor and reagent kit supplier.
|Nucleic Acid Extraction||Cell Viability & Cytotoxicity Assays||Cellular Change Assays|
When is the project deadline?
When does your automation project need to be finished? The answer will determine if you can take time to develop the workflow on your own or if you should pay an outside automation expert to support your method development. While robotics companies typically provide a small degree of training, you will typically need to pay for installation and development of new methods.
Developing robotic methods on your own can be very time consuming and resource intensive.
Where will the automated workflow be run?
The layout of the lab could put some constraints on your automaton system. You’ll need to take the space’s floorplan and electrical capacity and the system’s footprint into consideration.
- How much space will the robot require?
- Do you have tables or benches that can support the robot’s weight?
- Is there enough electrical power to support an instrument? Can you have the correct outlets installed.
- Can the floor support the instrument size? Some instruments can be very large. For example, the Roche Cobas 8800 system weighs 2405kg.
- Can the instrument be transported to the desired location?
Who will operate and maintain the workflow and instruments?
You should identify a lab scientist who will “own” the automated process as an Automation Lead. Consider investing in training from a liquid handling company for this user. Once you have established this institutional knowledge, modify your SOPs to maintain this knowledge, especially if you’re part of a small organization or academic core laboratory. Turnover can make managing the special training required for automation difficult.
How will the automation be implemented?
Those implementing new automated workflows should have experience in the chemistry and robotics and, ideally, the application. For example, a collaborative team made of a member of your lab, a robotics expert and a support specialist from your chemistry, robotics, and, ideally, the application.
Or you can partner with our Field Support Scientists, who have decades of experience automating our kit chemistries on a wide range of platforms and throughput scales. We’ll help you customize the scale, application, and chemistry exactly to your needs, regardless of platform.
What is your budget?
The price of an automation instrument can range from less than $10,000 for the simplest platforms to over $500,000 for the most complex. Customization, installation, warranties, training sessions, consumables and support fees can also drive up your project’s price tag. Knowing your budget can help you select an instrument and design an automated workflow that you can afford.
Automate Your Lab Worksheet
Plan your lab automation project with this worksheet.