OpenIDEO is an open innovation platform. Join our global community to solve big challenges for social good. Sign Up / Login or Learn more


Machine Learning, a toll-free helpline, mobile phone tech and a community-based parenting buddy system to improve ECD outcomes in LMICs.

Photo of Coenie Louw
7 1

Written by

Name or Organization

Gateway Health Institute NPC


South Africa. We operate at a national level, but we are based in Gauteng - Pretoria & Johannesburg.

What is your stage of development?

  • Early Stage Innovator, with at least one-year experience in ECD


  • Non-profit

What is the stage of your proposal?

  • Prototyping: I have done some small tests or experiments with prospective users to continue developing my idea.

Describe how your solution could be a game-changer for your selected Opportunity Area (600 characters)

In South Africa, the NDoH’s strategy of massification of Early Childhood Development (ECD) centres is ineffective. There is a clear gap in parental and healthcare worker support. The ECD span from the prenatal period to three-year old does not receive enough emphasis. Children in this first stage of ECD is often compromised when the mother (and the partner) is a teenager. There is little focus on supporting adolescent and young parents. SmartBaby aims to address this gap in health care provision to ensure optimal ECD for children 0 to 3 years old.

Select an Innovation Target

  • Service: A new or enhanced service that creates value for end beneficiaries.

Tell us more about your innovation (1500 characters)

SmartBaby will fill the gaps in parental and health worker support that is needed in South Africa to ensure babies develop to their full potential by focusing on the prenatal period through the first three years of a child’s life. SmartBaby uses a mobile app for learning, a toll-free helpline for counseling and support, and risk stratification (low, medium and high risk) through supervised machine learning. Specially designed care packages are developed according to risk segmentation of the mother, including parenting buddies to provide community-based support.

What problem are you aiming to solve? (3 sentences)

In South Africa, the NDoH’s strategy of massification of Early Childhood Development (ECD) centres is ineffective, the critical ECD span from the prenatal period to three-year old does not receive enough emphasis. There is a clear gap in parental and healthcare worker support - parents' questions go unanswered and front line health workers lack knowledge/skills Children in this first stage of ECD are compromised by the mothers' risk behaviour, socio-economic factors and lack of social support

Explain your idea (5000 characters)

Elements of SmartBaby 1. Supervised machine learning(SML) for risk segmentation. It requires two things to be successful: a rich dataset and a ‘measure of success’. A rich dataset will contain demographic, behavioural, geographic and psychographic information of the mother. The ‘measure of success’ is segmentation of mothers into risk groups and developing care/intervention packages based on the identified risk group(ing). 2. A Progressive Web App for evidence based information and social behaviour change communication. 3. A toll-free helpline to provide advice, support, counselling and linkage to care, including recommended care packages. 4.Community-based support by parenting buddies.

Who benefits? (1500 characters)

1. Parents and health workers from learning and support. Children from comprehensive ECD interventions. Communities from well developed children. 2. GHI has been working with women, adolescents and children in South Africa for 3 years.

What kind of impact will your idea have? (1500 characters)

Comprehensive ECD during the critical pre-natal and first 3 years of a child's life requires good nutrition, good health, a stimulating environment, and loving care in the early years of life, for the physical, mental and social development of a child. From better school performance to lower criminal behavior, the right combination of health care, adequate food, a learning environment and good parenting, installs qualities that result in more productive, more socially adapted, and in a general sense "better" children and adults, compared to when one of these factors is missing. Research provides evidence that children who experience stress during their early years are at greater risk of cognitive, behavioural and emotional difficulties later in life.

How does or how could your idea impact low-income children? (1500 characters)

SmartBaby is designed to impact the 30% of children in SA subjected to child income poverty, which is linked to adult unemployment(40%). It is therefore a free service using low-cost tech, cloud-based servers and SIP trunking that is sustainable.

Scale: Describe how your idea could reach a significant number of end-users. (1500 characters)

Mobile phone technology is ubiquitous. The system is designed to scale. Amazon Web Services servers and data centers can be scaled with the click of a button. Revenue streams include advertising and survey services, freemiums and consultancies.

Feasibility: Where are you with understanding the feasibility of your idea? Describe what you’ve done so far and your plans. (3000 characters)

GHI has in place an experienced team with a long term vision for our projects. The team includes Prof D Castelijn, Occupational Therapy; PhD, Prof A van Rensburg, PhD Industrial Engineering, Chantell Witten, MSC and PhD candidate, nutrition; Sara Nieuwoudt, MPH and PhD candidate, social behaviour change communication; Patricia Martin ECD expert; Dr Louw, MBcHB and HIV expert. Our project plan & milestones are solid, realistic and achievable. A sustainability plan with business model is in place.

Business Viability: How viable is your business model? (5000 characters)

Barriers In SA only 38% of people own a SmartPhone. Poverty and high unemployment is a barrier to parents accessing the SmartBaby Service Illiteracy 11 official languages in South Africa

HCD: How have you used human centered design to build or refine your concept? (5000 characters)

GHI focus on co-creation Community/Beneficiary participation in all aspects of a project - from design, development, testing, implementation to monitoring and feedback. Beneficiary participation is key to success and sustainability. First step is always focus group discussion for a needs assessment, in this case with a group of parents, including pregnant women from a disadvantages informal settlement near Johannesburg named Diepsloot. Why do they feel children from their community need, what do parents need, how involved are mothers, father, caregivers, the community in ECD, do they think ECD for children 0 to 3 years old is critical. What does ECD mean to them? nutrition, healthy child, stimulation. How do they understand CD milestones... Prototyping and testing SmartBaby with the same focus group for ease of use, appropriate content, willingness/ability to pay, affordability Iteration based on participatory play-testing

Tell us more about you (3000 characters)

GHI is a member of the South African Civil Society for Women's, Adolescents' and Children's Health(SACSoWACH) - a coalition of 25 organizations endorsed by NDoH. There are 3 working groups in the coalition - ECD is one of them, Nutrition another. My passion is finding solutions to challenges using technology - as a medical doctor I see potential in reaching vulnerable and remote groups with low-cost tech. Children and adolescents are my passion. Africa's future lies in well-developed children.

Do you have the people and partners you need to do what you’ve described? (600 characters)

Yes, we have a team and organizational partners with the expertise and experience to achieve success. Professor Daleen Castelijn, Department of Occupational Therapy, University of the Witwatersrand, Johannesburg. Patricia Martin, ECD expert, Orbis Africa/Advocacy Aid Sara Nieuwoudt, Social and Behaviour Change Communication Division, University of the Witwatersrand, Johannesburg. Porfessor Antonie van Rensburg, Department of Industrial Engineering, University of Stellenbosch. Chantell Witten, nutrition expert, Centre for Nutrition Excellence, North West University, Potschefstroom.

As you consider your next steps, what kinds of help could you use? Is there a type of expertise that would be most helpful? (1800 characters)

There is always room for more expertise and passion. We are also keen to share knowledge. Collaboration is key to success, not only among innovators but also with communities and organizations. We are open to mentors, collaborators and partners from any LMIC.

Would you like mentoring support?

  • Yes

Are you willing to share your email contact information submitted on OpenIDEO with Gary Community Investments?

  • Yes, share my contact information

[Optional] Biography: Upload your biography. Please include links to relevant information (portfolio, LinkedIn profile, organization website, etc).

Dr. Coenie Louw, MBChB holds a postgraduate Diploma in HIV Management. With 24 years’ experience in the medical field, he currently focuses on solving global health challenges through using innovative technology. As Executive Director of Gateway Health Institute he has won 5 global challenge awards over the last 2 years, as well as a People’s Choice Award from the UN’s Special Envoy for health MDGs and Malaria.

[Optional] Attachments: Please upload relevant attachments or graphics or show us how you prototyped.

Coming soon

[Optional] Video: You are invited to submit a 30-60 second video that introduces you and/or your team and your idea.

Coming soon

Mentorship: How was your idea supported? (5000 characters)

We were teamed up with Rob Weiss from We had a face to face meeting in Johannesburg ( yes, that was a bonus) and a follow up call ten days later. Our discussion focused mainly on sustainability and developing a business model that would ensure that revenue streams can be created to sustain SmartBaby beyond initial funding.

Updates: How has your idea changed or evolved throughout the Prize? What updates have you made to this submission? (1500 characters)

Decided on a SmartPhone app to test SmartBaby rather than USSD To put more focus on the concept that SmartBaby needs to be replicable and implementable in the USA To use our current platforms to generate income for SmartBaby Found a long term financial solution through our technology partner 4Sight Holdings - Wa'Ching


ECD from 0 to 3 years is a global challenge, including in the USA. We decided to focus on technology - a PWA and a toll-free helpline using Amazon Web Services that would be readily available, acceptable and affordable anywhere in the world PWA content can be easily adapted to the USA context

Scale: Describe how your idea could reach a significant number of end-users. (1500 characters)

PWA is digital and can be reached by any parent with internet access Cloud-based helplines, especially if toll-free can reach thousands of parents, eliminates issues around illiteracy and can be offered in a variety of languages. (As in SA with 11 official languages.) Cloud-based helpline peer counselors, peer educators and experts can be anywhere in country and does not require expensive call-center infrastructure and human resouces. The helpline can be manned by community volunteers on a roster.

Are you willing to share your email contact information submitted on OpenIDEO with Gary Community Investments?

  • Yes, share my contact information


Join the conversation:

Photo of Andreana Castellanos

Dear Doctor @Coenie Louw

My name is Andreana the founder of Afinidata - The artificial intelligence chatbot that helps parents boost their child’s potential we love to connect with you we are working in Latinamerica and U.S but we love to connect with other programs that are tech based and focus on scalability. Have a couple of questions in regard of how does the machine learning work with the different accents for the same language ?

Photo of Iliriana Kacaniku

Dear Doctor Coenie Louw 
Few days ago we shared with you the invitation to the two webinars in which you will have a chance to hear more about the Prize and ask any questions about the process. Knowing that the sponsor seeks solutions from all over the world that can be implemented in the United States, I think this is a great opportunity to hear any questions you may have on how to develop
your idea along this goal. And many other aspects of the Prize and human-centered design.

We’ll be hosting calls on
Tuesday, December 5, 4:30 - 5:30 PM PST ( and

Wednesday, December 6, 08:00 - 9:00 AM PST (

In case you cant make it, feel free to email us any question you may have at!, and we’ll do our best to answer it during the webinar.

Looking forward to seeing you next week,

Photo of Iliriana Kacaniku

Dear Coenie Louw 
I am so glad that you benefitted from both Webinar yesterday and the mentorship. Our team has worked hard to make sure that we engage expert mentors from all over and make sure that they guide our innovators the best.

Can't wait to read the updates to your idea as a result. Should you have any question before Feb 15, please reach out.

Best regards,

Photo of Iliriana Kacaniku

Dear Doctor Louw,
Thank you for submitting your idea to the challenge and educating me about impact problems in South Africa.

I read your description with great interest, and I like what you are trying to achieve. From my experience of managing numerous projects and recently completed MBA, I noticed 2 aspects of potential strengthening of your idea. Firstly, your idea may benefit from identifying few channels of reaching out to mothers. Mothers being one of the primary target beneficiaries, it will be very important to identify what approach would help you to onboard mothers as users, test the functionality of the solution, and develop further before full roll-out. Secondly, one of the key criteria of successful impact solutions is to engage the stakeholders throughout the process of impact design, development, implementation, and measurement. From your experience how feasible is this in South Africa, and which stage do you think would be most likely?

Best regards,

Photo of Coenie Louw

Dear Iliriana
Thank you for the comments.

When I first uploaded SmartBaby on this platform there was a limited number of characters (instead of 250 words it only allowed 250 characters). So at this stage the idea is very cryptic and short. We are working on expanding on the detail offline.
However, to answer your questions:
SmartBaby is all about the mother.
We have other platforms and projects to use as channels to reach mothers and get them on board. SmartParent - for which we recently won the Positive Action Challenge being one of them. We need a rich data set, and we plan to crowd source input from mothers, to enhance the development of the data set, the Machine learning algorithm, and the risk segmentation. We also have 72 Community health Workers and 300 Home Based Care Workers at our disposal, making it relatively easy to reach mothers.
All our projects are based on community participation, from development through to implementation, to results dissemination and feedback. Community participation is key to community ownership, and community ownership key to sustainability.
How do we involve the mothers - a community advisory board, focus group discussions, community dialogue and in this case, most importantly, crowd sourcing.
Here is a good example:
And yes, all of this works very well in South Africa and other LMIcs


Photo of Ashley Tillman

Hi Dr Louw! Great to see you in the Challenge. You mention that you've either prototyped or piloted this idea. Can you share a little bit about who you prototyped/piloted with and what you learned?

Photo of Coenie Louw

Hi Ashley

Gateway develops and prototypes our ideas with beneficiaries - in this case parents, caregivers and frontline health workers in primary health clinics - through focus group discussions (needs assessments) and then through testing (play testing, walk throughs) to allow for agile development and iteration.
Our expert panel continuously vet our content to ensure it is culturally appropriate (community advisory board) and evidence informed (academic experts).
Machine learning requires a rich data set, so the process is far from complete. But as you say iterate, iterate, iterate...