Striking the balance between automation and human skill on the seafood processing line

By Brian Teece, Regional Manager Scotland at Interfood Technology

Automation is being touted as the fix-all in food manufacturing. For seafood processors, it’s positioned as the answer for efficiency, productivity and labour shortages, among other things. But is it really enough? Can AI and automation really improve your operational performance enough, or do you still need human experience to make the difference between speed and quality? And what do you actually need to strike the balance, and make your automation investment truly work for your business?

Put repeatable work first

Automation provides value when it frees people to handle other tasks while delivering measurable results. That usually means focusing on repeatable tasks, such as weight grading and product movement, or processes that reduce manual handling. As long as any new equipment fits within the context of everyday factory operations, and it’s capable of handling the expected workload without creating extra work – consider cleaning and maintenance requirements and fault resolution – it can bring value by supporting your team. But that’s really where the focus needs to be: helping your team to do their jobs more easily and quickly, rather than attempting to replace them.

Frozen block breakdown is a good example. Done manually, it means repeated bandsaw passes and repositioning, with the operator’s hands close to the blade every time. An enclosed automatic cutting cycle, such as the Astech SRA 3.0 or SRA Compact for programmed thickness cuts and the CC-2 or CC-5 for longitudinal splitting, uses controlled feed and product gripping to reduce hands-on cutting, repositioning and blade exposure. The result is repeatable sections, subject to block dimensions and the cutting pattern required. HSE identifies repetitive production work and awkward lifting as causes of musculoskeletal injury, so there’s a safety case here as well as a throughput one.

Recognise the value of experience

The thing about seafood is that it’s rarely as predictable as the containers, trays and packaging used to handle it. Fish vary in size, shape and flesh condition, while previous handling can affect their appearance and structure. Fillets can also deform during movement, making precise handling more difficult than it would be with a rigid product. And that’s where you need experienced staff. They can judge whether a piece is sellable, whether tissue needs to be removed and how much, and what sort of treatment each item requires. This is surprisingly skilled work that technology hasn’t been able to effectively handle until recently. But with machine vision and adaptive cutting technology, there is now the opportunity to use automation to support experienced staff.

Portioning shows where that support pays. Operators typically judge steak thickness by eye, then weigh and correct. A weight-controlled portioner such as the Astech PW-Saw scans the product and cuts to programmed weight and thickness parameters, accommodating changing cross-sections along the fish. That reduces overweight giveaway and underweight portions, along with the manual weighing, trimming and rework that go with them. As an illustration, cutting average overweight by 5g across 10,000 portions a day retains 50kg of fish a day. At an assumed raw material cost of £6/kg, that’s £300 a day, or £75,000 over 250 production days, before operating costs and any change in offcuts or rejects. Accuracy still needs trials using your own fish at the intended processing temperature.

Human knowledge and experience are still integral to the process, but when you provide an experienced staff member with technology that has been tested against your actual range of products, you help them to work more productively. Just remember to trial the tech before making that decision, measuring both defects that go undetected and acceptable products wrongly rejected, to assess how that could impact your bottom line, and whether these mistakes could be monitored and addressed with human oversight.

Measure performance across different production runs

Of course, automation becomes a much more viable option when your production line handles a single product, or long runs of similar products. If your primary task is processing salmon steaks, it’s far simpler to introduce automation than if you’re regularly managing short runs of tuna, mackerel, herring and trout, as well as salmon, each with different cutting specifications and packing requirements.

To gauge the effectiveness of automation across your production, you need to take into account how much variation the equipment will face. Not just now, but during seasonal changes, which are likely to impact your product range and volumes. Also consider whether you’re likely to require future recipe changes, and how easy that might be to initiate with the new equipment, as some systems are far more involved than others.

These are the things that your trials should reflect. While running speed matters, it’s not the only priority. Setup, ease of cleaning, required time offline and saleable yield all need to be measured, if you’re going to ascertain the actual value automated equipment can bring. Account for sawdust and offcuts too when assessing any claimed yield improvement.

Find the constraint before adding capacity

Perhaps most importantly, before investing in new equipment, you need to establish where production is actually losing time. It’s tempting with AI and automation to simply choose a product because it sounds like it could be useful, but that often results in targeting processes that don’t need to be improved. So, talk to your team – managers and operators – to work out where there’s friction; where delays often happen, where errors occur. That’s where you need to focus.

Simply adding an automated portioning machine or robotic loader can often cause more problems than they solve if your team can’t pack quickly enough to keep pace. Sometimes, it can be worth doing, as part of a complete workflow redesign, but that’s only a sensible option if that was your intention all along. Otherwise, it simply introduces another bottleneck for your team to work around. Infeed matters here as well. An automated loader will spend much of its time waiting if fillets arrive overlapping or inconsistently positioned, so improving the transfer or the product presentation is often a better first step than the machine itself.

Make people part of the solution

Good automation doesn’t have to mean removing people from production. Fish Focus’s coverage of onshore processing automation shows how investment introduced in stages around existing factory layouts can be of the most benefit. This incremental approach can allow you to improve selected operations while retaining valuable skills elsewhere. So, automation tools could handle routine portioning and product movement, while your employees focus on quality checks, finishing and unusual cases. A conveyor between stations can remove the repeated carrying that pulls experienced people away from quality-critical tasks. Vision systems might identify potential defects for trained staff members to review. You just need an established workflow to deal with it.

So, jobs may change, and certain manual tasks may completely disappear, but you’ll still need people for the things that shouldn’t be left to technology: loading, supervision, cleaning, maintenance and exception handling. And that also needs to be budgeted for when you’re making investment decisions.

Successful automation starts with understanding the process, not selecting the technology. For seafood processors, the aim is to introduce consistency where it adds value while retaining human adaptability where it matters most. Getting that balance right isn’t always easy. But it’s always worth the effort when it works.