Abstract
The South African traditional beer (umqombothi) has been an isolated product for decades. This has been due to its production complexities, longer fermentation times, and quality inconsistencies. This presents a challenge to brewmasters with the desire to optimise the process, reduce cost and improve quality. The fourth industrial revolution (4IR) and its emerging technologies are increasingly becoming useful tools in the food and beverage industry. Specifically, beverage bioprocess development and optimisation are moving in a new direction. This involves the use of combinational experimental methods, replacing empirical strategies. As a result, classic bioprocess optimisation methods such as response surface methodology (RSM) need to be reinforced and validated further using intelligent systems. In bioprocessing, artificial neural networks (ANN) are a powerful nonlinear multivariate tool with excellent generalisation, prediction, and validation capabilities. Response surface methodology and ANN were thus used to develop and optimise umqombothi’s bioprocess. The coefficient of determination (R2) for all the parameters was closer to 1, indicating that RSM was an effective method for optimising umqombothi’s bioprocessing parameters. The investigated parameters during optimisation were alcohol content, total soluble solids (TSS), and pH, with R2 values of 0.94, 0.93, and 0.99 respectively. Similarly, the R2 for alcohol content, TSS, and viscosity in the constructed ANN was 0.96, 0.96, and 0.92 respectively. As result, a good correlation between the experimental and predicted values showed that a coupled approach had a positive impact on the bioprocess and the final product. The optimal processing conditions were cooking the mixed ingredients for 1.10 h at 95 ℃, while the optimal fermentation conditions were 29.30 ℃ for 25.90 h. As hypothesised, these conditions positively influenced the nutritional composition of the final product, as reflected in the beer’s proximate compositions, minerals, amino acids, B-group vitamins, and sugar compounds were determined. In comparison to the other samples (i.e., customary beer brew (CB) and mixed raw ingredients (RI)), the optimised beer brew (OPB) had more energy (165 kcal/100 g), crude protein (8.57%), and ash content (1.01%). All the samples were relatively high in essential macrominerals and low on potentially toxic elements. Glutamic acid was the highest detected amino acid, with concentrations of 1.49 g/100 g, 1.54 g/100 g, and 1.62 g/100 g in the RI, CB, and OPB, respectively. The OPB contained a higher concentration of the two forms of vitamin B3, nicotinamide (0.16 μg/g) and nicotinic acid (0.74 μg/g) in comparison to the CB. The concentration of mannitol was 0.42 mg/g, 0.23 mg/g, and 1.53 mg/g in the RI, CB, and OPB, respectively, and OPB had the highest total amino The South African traditional beer (umqombothi) has been an isolated product for decades. This has been due to its production complexities, longer fermentation times, and quality inconsistencies. This presents a challenge to brewmasters with the desire to optimise the process, reduce cost and improve quality. The fourth industrial revolution (4IR) and its emerging technologies are increasingly becoming useful tools in the food and beverage industry. Specifically, beverage bioprocess development and optimisation are moving in a new direction. This involves the use of combinational experimental methods, replacing empirical strategies. As a result, classic bioprocess optimisation methods such as response surface methodology (RSM) need to be reinforced and validated further using intelligent systems. In bioprocessing, artificial neural networks (ANN) are a powerful nonlinear multivariate tool with excellent generalisation, prediction, and validation capabilities. Response surface methodology and ANN were thus used to develop and optimise umqombothi’s bioprocess. The coefficient of determination (R2) for all the parameters was closer to 1, indicating that RSM was an effective method for optimising umqombothi’s bioprocessing parameters. The investigated parameters during optimisation were alcohol content, total soluble solids (TSS), and pH, with R2 values of 0.94, 0.93, and 0.99 respectively. Similarly, the R2 for alcohol content, TSS, and viscosity in the constructed ANN was 0.96, 0.96, and 0.92 respectively. As result, a good correlation between the experimental and predicted values showed that a coupled approach had a positive impact on the bioprocess and the final product. The optimal processing conditions were cooking the mixed ingredients for 1.10 h at 95 ℃, while the optimal fermentation conditions were 29.30 ℃ for 25.90 h. As hypothesised, these conditions positively influenced the nutritional composition of the final product, as reflected in the beer’s proximate compositions, minerals, amino acids, B-group vitamins, and sugar compounds were determined. In comparison to the other samples (i.e., customary beer brew (CB) and mixed raw ingredients (RI)), the optimised beer brew (OPB) had more energy (165 kcal/100 g), crude protein (8.57%), and ash content (1.01%). All the samples were relatively high in essential macrominerals and low on potentially toxic elements. Glutamic acid was the highest detected amino acid, with concentrations of 1.49 g/100 g, 1.54 g/100 g, and 1.62 g/100 g in the RI, CB, and OPB, respectively. The OPB contained a higher concentration of the two forms of vitamin B3, nicotinamide (0.16 μg/g) and nicotinic acid (0.74 μg/g) in comparison to the CB. The concentration of mannitol was 0.42 mg/g, 0.23 mg/g, and 1.53 mg/g in the RI, CB, and OPB, respectively, and OPB had the highest total aminoThe South African traditional beer (umqombothi) has been an isolated product for decades. This has been due to its production complexities, longer fermentation times, and quality inconsistencies. This presents a challenge to brewmasters with the desire to optimise the process, reduce cost and improve quality. The fourth industrial revolution (4IR) and its emerging technologies are increasingly becoming useful tools in the food and beverage industry. Specifically, beverage bioprocess development and optimisation are moving in a new direction. This involves the use of combinational experimental methods, replacing empirical strategies. As a result, classic bioprocess optimisation methods such as response surface methodology (RSM) need to be reinforced and validated further using intelligent systems. In bioprocessing, artificial neural networks (ANN) are a powerful nonlinear multivariate tool with excellent generalisation, prediction, and validation capabilities. Response surface methodology and ANN were thus used to develop and optimise umqombothi’s bioprocess. The coefficient of determination (R2) for all the parameters was closer to 1, indicating that RSM was an effective method for optimising umqombothi’s bioprocessing parameters. The investigated parameters during optimisation were alcohol content, total soluble solids (TSS), and pH, with R2 values of 0.94, 0.93, and 0.99 respectively. Similarly, the R2 for alcohol content, TSS, and viscosity in the constructed ANN was 0.96, 0.96, and 0.92 respectively. As result, a good correlation between the experimental and predicted values showed that a coupled approach had a positive impact on the bioprocess and the final product. The optimal processing conditions were cooking the mixed ingredients for 1.10 h at 95 ℃, while the optimal fermentation conditions were 29.30 ℃ for 25.90 h. As hypothesised, these conditions positively influenced the nutritional composition of the final product, as reflected in the beer’s proximate compositions, minerals, amino acids, B-group vitamins, and sugar compounds were determined. In comparison to the other samples (i.e., customary beer brew (CB) and mixed raw ingredients (RI)), the optimised beer brew (OPB) had more energy (165 kcal/100 g), crude protein (8.57%), and ash content (1.01%). All the samples were relatively high in essential macrominerals and low on potentially toxic elements. Glutamic acid was the highest detected amino acid, with concentrations of 1.49 g/100 g, 1.54 g/100 g, and 1.62 g/100 g in the RI, CB, and OPB, respectively. The OPB contained a higher concentration of the two forms of vitamin B3, nicotinamide (0.16 μg/g) and nicotinic acid (0.74 μg/g) in comparison to the CB. The concentration of mannitol was 0.42 mg/g, 0.23 mg/g, and 1.53 mg/g in the RI, CB, and OPB, respectively, and OPB had the highest total amino The South African traditional beer (umqombothi) has been an isolated product for decades. This has been due to its production complexities, longer fermentation times, and quality inconsistencies. This presents a challenge to brewmasters with the desire to optimise the process, reduce cost and improve quality. The fourth industrial revolution (4IR) and its emerging technologies are increasingly becoming useful tools in the food and beverage industry. Specifically, beverage bioprocess development and optimisation are moving in a new direction. This involves the use of combinational experimental methods, replacing empirical strategies. As a result, classic bioprocess optimisation methods such as response surface methodology (RSM) need to be reinforced and validated further using intelligent systems. In bioprocessing, artificial neural networks (ANN) are a powerful nonlinear multivariate tool with excellent generalisation, prediction, and validation capabilities. Response surface methodology and ANN were thus used to develop and optimise umqombothi’s bioprocess. The coefficient of determination (R2) for all the parameters was closer to 1, indicating that RSM was an effective method for optimising umqombothi’s bioprocessing parameters. The investigated parameters during optimisation were alcohol content, total soluble solids (TSS), and pH, with R2 values of 0.94, 0.93, and 0.99 respectively. Similarly, the R2 for alcohol content, TSS, and viscosity in the constructed ANN was 0.96, 0.96, and 0.92 respectively. As result, a good correlation between the experimental and predicted values showed that a coupled approach had a positive impact on the bioprocess and the final product. The optimal processing conditions were cooking the mixed ingredients for 1.10 h at 95 ℃, while the optimal fermentation conditions were 29.30 ℃ for 25.90 h. As hypothesised, these conditions positively influenced the nutritional composition of the final product, as reflected in the beer’s proximate compositions, minerals, amino acids, B-group vitamins, and sugar compounds were determined. In comparison to the other samples (i.e., customary beer brew (CB) and mixed raw ingredients (RI)), the optimised beer brew (OPB) had more energy (165 kcal/100 g), crude protein (8.57%), and ash content (1.01%). All the samples were relatively high in essential macrominerals and low on potentially toxic elements. Glutamic acid was the highest detected amino acid, with concentrations of 1.49 g/100 g, 1.54 g/100 g, and 1.62 g/100 g in the RI, CB, and OPB, respectively. The OPB contained a higher concentration of the two forms of vitamin B3, nicotinamide (0.16 μg/g) and nicotinic acid (0.74 μg/g) in comparison to the CB. The concentration of mannitol was 0.42 mg/g, 0.23 mg/g, and 1.53 mg/g in the RI, CB, and OPB, respectively, and OPB had the highest total aminoacid levels. Overall, OPB displayed a desirable nutritional profile compared to the CB. Thus, this study highlights the effectiveness of a coupled approach in developing a standard and optimised production process for a nutritious, high-quality beer.
M.Sc. (Biotechnology)