Page 30 - FoodFocusThailand No.242 June 2026

SPECIAL
SPECIAL FOCUS FOCUS

SOFT ROBOTICS:

FROM ROBOTIC ARMS TO SMART ROBOTS

IN THE MODERN FOOD INDUSTRY

The food industry is shifting from traditional automation to smart factories that can sense, decide, and adapt
in real time. This change is driven by rising labor costs, stricter food safety requirements , decarbonization
1
pressures, and higher consumer expectations for quality. The key technologies behind this transformation
are integrating AI with robotics and production line sensors (IIoT). Consequently, robots are evolving from
mechanical handling arms into intelligent, learning machines.

In the future, humans and robots will increasingly work production parameters and continuously learns from data.
together. Robots will handle repetitive, heavy, or hazardous This architecture requires a high-frequency, low-latency data
tasks, while humans will focus on decision-making, quality infrastructure (PLC/SCADA) and seamless OT/IT integration,
supervision, and complex problem-solving. Collaborative ultimately minimizing waste, downtime, and variation while
robot (Cobot), will therefore play an important role in improving boosting traceability.
safety, speed, and efficiency in food production.
Soft Robotics: Precision and Gentle Handling for
From End-of-Line Inspection to Food Applications
In-Line Quality Control Major challenge in food production is handling products that
For traditional food processing plants, production process are fragile, slippery, soft, or irregularly shaped. Examples
control still relies on maintaining key variables within specified include ripe fruit, bread, fresh meat, and products with uneven
ranges. These variables include temperature, pressure, flow surfaces. Conventional rigid grippers may damage these items
rate, heating time, and humidity, which are managed through because they apply pressure unevenly and cannot easily adapt
PID controllers and static production recipes. Quality is then to different shapes.
inspected again at the final stage through random sampling Soft Robotics addresses this problem by using flexible
to measure values such as a , pH, color, viscosity, crispness, materials and adaptive structures. These may include
w
sugar content, and microbiological contamination. Elastomers, Pneumatic Networks, or collapsible designs that
Predictive Quality Control (PQC) shifts decision-making distribute contact force more gently, reduce localized pressure,
to “in-line processing,” using data-driven models to predict and adapt better to the shapes of food items . Modern Soft
2,3
quality in real time. Systems track temperature, pressure, Robotics 2.0 requires three key capabilities.
motor torque, weight, and Near-Infrared (NIR) data for raw 1) Materials and designs must resist moisture, fats, cleaning
material composition, alongside cameras for appearance and agents, temperature changes, as well as contaminated areas
acoustic sensors for machinery health. and easy to clean according to hygienic design principles.
When historical batch data is processed using Machine 2) Precise grasping through Tactile Sensing allows robots
Learning (regression, random forest, neural networks), the to perceive contact forces, detect slippage, and adjust gripping
system accurately links process profiles to final quality. This pressure in real time to reduce food damage. STFT/DWT signal
enables real-time predictions and immediate live adjustments. analysis, and Machine Learning combined with camera data
For near-zero-waste manufacturing, a closed-loop and joint torque can perceive the physical characteristics of
control system is essential. It automatically adjusts food much more accurately than traditional robots. 6

30 FOOD FOCUS THAILAND JUN 2026

22/5/2569 BE 16:56
27-31_Special Focus_Ravis.indd 30
27-31_Special Focus_Ravis.indd 30 22/5/2569 BE 16:56