Skip to Main Content

Vaccine injection robot

Innovative idea!

Not really just age related, but I'm wondering if anyone has come across a robot that could be used to inject Covid 19 vaccines?  I've found automatic injection systems for cattle but nothing so far for people.  Given the need to inject milions of people and the conventional solution of training people specifically to do this I'd argue that an automated system should be safer and more consistent.  And probably several stations working side by side, supervised by one clinician could be much more productive.

As an engineer, I would imagine that the key to giving a reasonably pain free injection is (1) selection of point and angle of entry 2) a quick push to pierce the skin (3) a steady rate of injection using consistent force on the piston and (4) smooth extraction.  I'd welcome feedback on this.  Is the injection into the arm muscle and what sort forces, depth of penetration and speed of injection are typical?

This does not need to be a multi axis device so could be relatively inexpensive to produce.  I've already come across a plan to prepare syringes pre-filled rather than vials which seems like a good idea to improve reliability and productivity.

You will need to login to post a comment

Miroslava Katsur Dec 13, 2020

Dear Adrian, thank you for sharing this great idea! I have not heard of such robot before, but after thinking about it I realise that such robot may be used around the UK (perhaps, even driven across country to remote areas). I see the advantages of it - people can come when it is convenient for them and use it, plus healthcare workers and pharmacy staff can stay safer and use their time for other things instead. I found that unmanned vaccination booths are already being used in the central Henan province, so this is something that can be trialled in the UK as well. Also, I read now that "France-based NextMotion, which focuses on connected aesthetic activity using imaging and other digital tools, has been trialing LENA, a light-enabled neuro-robotic arm" which is maybe the first robot which administers injections. LENA is, however, is probably a very expensive robot, so I really hope that, if necessary, we can produce inexpensive robots fast to vaccinate British population asap.

If to consider older generation, I think that the injection should be done for a sitted person (just to ensure they are not moving much) with the possibility to fix one arm in place. Perhaps, there should be adjustments made to allow people with disabilities to be able to use such robot.

Also, such robot can be used not just for COVID-19 vaccines. There are other vaccines which are recommended for older adults, such as influenza (1 per year), tetanus booster (1 per 10 years), Varicella (2 doses), Zoster (2 doses) and others.

I hope other people can contribute to the discussion of this topic. Thank you, Adrian, for sharing this idea!

Users tagged:

Reply 0

Adrian Girling Dec 14, 2020

Dear Miroslava, thanks for finding that this is already happening in China and also with LENA in France.

I very much agree, it needs to be a relatively inexpensive solution. The China version, see , uses a "robotic stirrup" to rest the arm whilst the injector arm looks like it has two axes of movement.

I've tried to describe what I have in mind and realise, as usual, that it's better to sketch something which I will find time to do!

I believe that clinicians taking vaccine from a vial would turn it needle upwards to let the air come to the top, tap it to make sure there are no air bubbles and then press the plunger a little to make a small amount of liquid come out, ensuring all the air is out. I imagine this needs to happen which makes things a bit more complicated but I'd welcome inputs as to how critical this is. I can imagine it's more critical for an injection is into a vein and if not done it would be more painful into a muscle but how important is it that this is done expertly? The human ability to do this reliably and to see if it has worked propoerly would be hard to beat!

Reply 2

Miroslava Katsur Dec 14, 2020

Hi Adrian! Thanks for offering to sketch your idea - I'm looking forward to seeing it!

I saw how COVID-19 vaccine is injected and some others, and it seems that the needle needs to go into the muscle. I think it is also safer, easier and cheaper to manufacture a machine which inject into the muscle rather than into a vein (I may be wrong - I'm not an engineer!). I found papers though which say that it is important to give vaccines into the muscles. This, researchers argue, minimises adverse effects (e.g. higher immune response to vaccine).

Bubble removal can be avoided by prefilling the syringes with vaccine, as you already mentioned. I routinely use syringes for other purposes (studying nanoparticles) and I try to remove even tiniest bubbles from them. I know that if I (a) remove all visible air bubbles and (b) do not shake liquid/syringe (for example, vaccine or syringe with vaccine), then there may be only a few bubbles in nanometer size, which I do not expect to cause problems if injected into the human muscle. Cells are usually bigger than those bubbles which I get. Also, all blood vessels except for capillaries are bigger than these air bubbles.

Vacuum can be used to remove the bubbles from the vaccine as well. This is something I do as well. It does not take much time, nor sophisticated equipment.

If I think of something else to add to this discussion, I'll write later on!

Users tagged:

Reply 0

Adrian Girling Dec 14, 2020

Hi Miroslava, this is really helpful. If pre-filled syringes are so constructed as to avoid any air entrapment even at the end of the needle this has to be the best solution. There is no point in trying to automate this at the point of use when the preparation of pre-filled syringes is sure to be fully automated.

Muscle is much easier to find than a vein! I'm guessing, the best place to inject is in the arm at the point opposite the armpit, maybe a little lower.

I'll get my thinking cap on. Further inputs most welcome.

Reply 1

Miroslava Katsur Dec 14, 2020

Hi Adrian!
I found that the muscle recommended for the use is the deltoid.
I guess as long as it gets into the muscle, it should be fine. Perhaps, the thickest portion of the muscle can be targeted.
Thanks for your reply!

Users tagged:

Reply 0

Adrian Girling Dec 14, 2020

Useful link, many thanks indeed

Reply 0

Miroslava Katsur Dec 13, 2020

Status label added: Innovative idea!

Reply 0

Georgie Cade Dec 14, 2020

This is really interesting, and it would be good to hear some more thoughts around this.

It sounds like it would definitely be more efficient, but I wonder if one element that may be a barrier to this is actually public perception of injections? Some people may be feeling nervous about being injected or have had scary misinformation about vaccinations, in this instance having a machine injecting them may escalate their worries or remove some of the 'human' comforting side of the interaction Perhaps having the clinician supervising would get rid of this fear though?

With engineering I often find there is some tension between what is technically possible and what people feel comfortable with using or interacting with out in the 'real' world. Have you come across similar tensions in your experiences Adrian?

Reply 1

Adrian Girling Dec 14, 2020

Thanks Georgie for your enthusiastic reply.

In terms of public perception logic doesn't always win, but I'd argue that I'd rather be innoculated by a precise robot than someone who has recently been trained and whose method may not the so consistent. Am I allowed to say, I've had some injections which hurt a lot, and some that didn't hurt at all. I do think having a clinician supervising a number of stations is a good idea anyway - a nurse to instruct and reassure users as they arrive.

I agree, people are naturally skeptical of something with which they are not familiar - a natural survival instinct. On the other hand, humans seem to be incredibly good at adapting to new circumstances - nobody would have guessed how succesful most people have adapted to working from home this year. I can't imagine how we'd cope today without the internet and mobile phones...

Reply 1

Neel Desai Jan 4, 2021

Dear Adrian, I have no doubts that the use of a robot/other mechanical rig for vaccinations would improve the efficiency of delivery, not just for the COVID-19 vaccines but also others (such as the flu vaccination) since face-to-face contact can be limited.

As a pharmacist, I've been trained to vaccinate the general population and before the injection process there are many considerations that must be taken into account for. These include, injection site (left/right deltoid in the upper arm normally), injection angle (most often 90 degrees, but for some patients, 45 degrees) and speed of injection (usually 5 seconds per mL of fluid) to name a few. You have of course alerted to these in your step-by-step process above.

Whilst there is much simplicity to the idea, in comparison to livestock, which are often genetically 'produced' to be near enough identical, that is of course not true for humans! If such an idea could come to fruition, I imagine there would need to be an artificial intelligence able of 'sizing up' individuals before administration.

Such a machine would need to be supervised by a healthcare professional on the off chance that anaphylaxis occurs. This may also help to comfort patients.

Overall, there are 'real-world' elements that I think are often challenging to address. But as I mentioned earlier, this could be the way forward if engineering and healthcare teams work together!

Would love to hear your thoughts on AI-driven robots to help vaccination from an engineer design rig. Thanks again for a thought-provoking idea Adrian!

Users tagged:

Reply 0

Adrian Girling Jan 5, 2021

Dear Neel,

Many thanks for your input. I’m slowly gathering key information to help me wite a proper specification requirement. I do need to get the core variables right – knowing that 5 seconds per ml is usual is helpful. It would be useful to know the typical range of ml for covid, flu and other vaccines and the corresponding internal diameters of syringes. For an automated system I’d feel most comfortable with a small diameter syringe so that fluid volume with time is easier to control.

Interesting that the angle is typically 90 or 45 degrees – is this vaccine or patient related?

I completely agree, it’s important that a healthcare professional is in attendance – the aim is to improve their productivity. I would imagine that the process includes an interview with a healthcare professional to establish whether the patient has allergic reactions and/or any underlying health conditions or is taking medicines.

Presumably BMI (possibly estimated by the healthcare professional) is important – the circumference of the arm at the injection point probably determines the ideal depth to push the needle – through the fat, into the muscle, not as far as the bone! And does the injected volume vary according to BMI?

The overall complexity of automation now suggests a step wise approach to such a development. Although a completely automated injection station may be possible eventually, I’m leaning towards a product which the operator holds like a pistol where a crook on an extension arm is fitted under the armpit, the syringe is moved forward close to the arm, stopped by a locating plate with a hole in it pressing against the arm with the needle maybe 10mm behind it before a leadscrew with motor pushes the needle into the arm and then pushes the piston at pre-set speeds. The distance between the armpit crook and the locating plate should provide information about the ideal depth of needle insertion.

Loading a prepared syringe into the ‘gun’ should be a simple click. Do syringes get re-used with just a change of needle or is the whole assembly disposed of after an injection?

For mass injections, this would make it quicker and easier to train operators to administer injections safely and consistently. Indeed, a completely automated cubicle may need someone in attendance to explain how it works and it would probably be slower as patients are shown how to use it rather than just being told to sit still and close their eyes!

I’m a newcomer to the concept of AI but my thinking is that it relies on having a body of data to start with. Presumably basic allergy, underlying health, age, BMI and other data are being collected at the time as well initial reactions to the injection and hopefully any subsequent data.

Whilst all the variables like the depth of the needle and speed of injection of fluid would be set up initially according to best practice, AI could quickly look for correlations if all the data were brought together to suggest adjustments.

Just a footnote. Such a shame that Powderject doesn’t work, it would make the whole process so much quicker and simpler. I believe it fails because the fluid doesn’t go far enough to get into the muscle but I wonder if it deserves further investigation…


Reply 0