top of page
TAGE Labs Logo | Custom Research Equipment and Automation Solutions

TAGE LABS

Automation On A Research Budget

Updated: Jul 24

There is never enough
There is never enough

We all know that manual tasks equal labor hours. It is estimated that up to 30% of a lab’s budget can be consumed by repetitive tasks like pipetting, data entry, animal feeding, environmental controls or sample tracking. This does not include the potential for human error, missed observations, breaks or misses in scheduled activities and the list goes on. What if you could reclaim those funds for high-value work?


Even common tasks like data entry can have enormous cost impacts. Either directly through time taken, or indirectly via human error. These can be from random observational errors to more systemic issues related to environmental, observational or instrumental causes. Finally, straight forward mistakes are made. All of these can invalidate data or invoke repeating the work, all of which comes at significant expense.


Financial Cost

Looking at some cost examples:

  • Labor Costs: Calculate hourly wages spent on manual tasks

  • Error Tax - those mistakes cost real money and add up

  • Repeat Cost:  Having to redo experiments due to errors or bad data.


Automation is not just a tool, it really does multiply the cost savings, directly or indirectly.

  • Low-Cost Examples:

    • "Open-source scripts automate image analysis (no coding skills needed)."

    • "Low cost robotic solutions to replace hours of manual tasks"

    • "Cloud-based tools cut software/hardware  maintenance costs."

  • Look at long term ROI: A $3,000 automation solution could save $15,000 over the next year


Automation does not have to be only for well funded labs. There are low cost options from commercial solutions to home grown systems.


There will always be hesitation in adopting an automated solution. Let’s examine a few of the common pushbacks:


  1. "We can’t afford it." 

It is true that some solutions can be very expensive, yet there are alternatives. 

  • Open-source software is a popular option. 

  • The prevalence and affordability of 3D printing is also driving down costs. 

  • Low costs micro-controllers that are growing in computing power are expanding budget friendly options. 

  • Many solutions may offer tiered pricing models too.

  • "It’s too complex."

    Perhaps for a home-grown solution the automation may be too complex, requiring expertise to implement that is not available. Other affordable solutions are designed with the user in mind, to simplify and streamline operational tasks. Growth of the no/low-code software environments also helps simplify the development of in-house solutions. The growing availability of “off-the-shelf” sensors and other hardware options simplify what used to be complex solutions.

  • "Our workflow is too unique."

This is a very common hesitation. Truth be told, while the details of the experiment setup may be unique, a more generic or flexible automation solution could still be of benefit. If 85-90% of the work could be solved through automation is that not still worth the cost savings? Perhaps modular systems can be applied to tackle specific aspects of the experiment setup. Identify the highest cost components of the workflow and start there. Finding a solution from another task may be close enough to achieve most of what is needed in your experiment.


Data Quality

Research projects are heavily dependent on quality, integrity and reproducibility of data. Besides the indirect cost of human error, the data can become unreliable at best and unusable at worst. Peer reviews of published articles are strengthened by consistent and eligible data collection techniques. Ability to reproduce the experiment and achieve the same results and validate the research is strengthened by the ability to replicate the data collection process.


Automation is the way to standardize these experiments. Firstly in terms of input, or how the experiment is controlled, managed and regulated. Secondly, in the output, via data collection and management. This standardization across the experiment provides the reproducibility, quality and confidence in the final data sets.


Lost Time

Time is probably the most valuable resource for any researcher. That time should be spent judiciously. Asking the question :am I better off spending 3 hours a day counting samples or developing a new study technique?” The answer is somewhat obvious. Having highly skilled researchers spending precious time on mundane and repetitive tasks is not a high value proposition, even from a simple ROI standpoint.


Automation not only speeds up repetitive tasks but can:

  • Operate around the clock

  • Prevent redoing work due to mistakes

  • Free up time for more important tasks


All or Nothing

This can be a limiting approach. Believing that if an automation solution does not achieve 100% of eliminating human interaction, it's not worth investigating, may be limiting your opportunities. Oftentimes an “acceptable” approach may only target 50% or 80% of the complete project. That is still saving time and addressing the other problem areas.


Actions you can take now:

  • Try to automate one repetitive task this month - Analyze manual tasks

  • Identify the “low hanging fruit” - most laborious tasks that could be automated

  • Identify what should and could be automated

  • Explore free or open source solutions

  • Identify low-cost commercial options

  • Estimate costs to implement

  • Perform simple cost benefit analysis - Don’t ignore indirect costs

  • Shoot for 80% - not all or nothing


Summary

These common research problems can easily be addressed with some form of automation


Problem

Impact

Automation Solution

Cost

Direct and Indirect Budget Impact

Reduce human error, training, resource costs

Time

Ineffective use of time

Delegate tasks to automation 24x7Save time from not repeating and freeing up high skilled resources

Integrity

Low confidence in manual process and data

Reproducible and quality data collection



Conclusion

Automation does not mean retrofitting an entire lab or investing in a million dollar robotic system. It can be at the simplest level and expand over time. The cost saving is always the first benefit one thinks of, but data integrity and confidence is arguably a lot more important to credible research.


Automation solutions do not need to be expensive. There are ways to automate without destroying the budget. Sometimes the simple solution can have significant benefits without having to solve 100% of the problems..Start now with simple tasks and small steps.


--------------------------------------------------------------------------------------------------------------------

References:

Kimberly A. Barchard, Larry A. Pace,

Preventing human error: The impact of data entry methods on data accuracy and statistical results,Computers in Human Behavior,Volume 27, Issue 5,2011,(https://www.sciencedirect.com/science/article/pii/S0747563211000707)




 
 
 

Comments

bottom of page